Generated: 2026-03-04 Purpose: single-file internal read for Tyler.
This document compiles the current handoff, 72h integration record, core THARP working docs, cross-session intelligence synthesis, and ops status into one continuous read.
This package is the internal-share boundary for recent THARP work, meeting intelligence imports, and gateway operational status.
core/: planning docs, decision memos, call sheets, and
prompt used for next-cycle execution.meeting-intel/: synthesized cross-meeting intelligence
plus call-level summaries/transcripts.ops/: sanitized operational summaries and redacted
event logs.Primary source was synced from macstudio over Tailscale
in the prior 72h pull workflow and consolidated into Wake docs.
Generated: 2026-03-03 21:21:43 MST Source host: macstudio (Tailscale) Window: last 72 hours
Date: 2026-03-03
Prepared for: Patrick Smith
Purpose: Consolidate all THARP call intelligence and define the exact
path for tomorrow's project-management discovery session.
This synthesis uses:
2026-03-02 13:53, 65
min)2026-03-03 10:55,
55 min)2026-03-03 16:21, 46
min)Estimate Sync first then
Scheduling Intelligence, orDesign/Pricing lane:
Production/Scheduling lane:
Scoring: Severity, Universality,
30-day Buildability on 1-5.
Designer Estimate Sync (CV -> WIAT) with validation
Scheduling Backlog Autopilot + Live Exception Board
Quote-to-Actual Drift Monitor
Custom-vs-Repeat Job Routing Intelligence
Treat tomorrow (Chris, project management) as the decision gate for sequencing:
Date: March 4, 2026
Audience: Santiago, Garth
From: Patrick Smith / Candlefish.ai
Subject: First build recommendation from THARP discovery interviews
Approve a 30-day pilot for a designer-stage Cabinet Vision -> WIAT/Sightline estimate sync + validation layer as Build 1.
Across THARP discovery calls, the highest-confidence repeated pain is manual re-entry between systems at the designer estimating stage. The issue is not design quality; it is data handoff reliability.
Observed impact:
Parallel production findings show additional burden:
These two lanes connect operationally, but the estimating sync lane is the most buildable and fastest to measurable ROI in 30 days.
Deliver a scoped sync workflow that:
Design principle: remove repetitive re-entry without forcing process replacement.
We will lock a tighter ROI range after PM call data confirms customer-facing rework and exception rates.
Pilot success metrics:
Following today’s PM session, we will deliver final Build 1 scope, risk controls, and timeline for approval, then start implementation immediately on sign-off.
Purpose: Run a decision-grade project-management discovery call that resolves build sequencing for the next 30-day pilot.
Call owner: Patrick Smith
Planned duration: 55-60 minutes
Decision target by end of call: choose Build 1 scope and
lock post-call data package.
Script: "Today is strictly workflow mapping and pain quantification. We are not replacing your systems. We are identifying one narrow fix that removes repeated manual work and reduces customer risk."
Ask in order:
Capture:
% status chasing vs
% core PM.Ask in order: 4. "What are the most common customer questions that are slowest to answer?" 5. "Where do delivery/install surprises typically originate?" 6. "Which updates can you send confidently without manual cross-checking?" 7. "What internal mismatch most often becomes customer-visible?"
Capture:
Ask in order: 8. "Which fields in WIAT/Sightline do you trust, and which do you ignore?" 9. "When data conflicts, what source becomes final truth?" 10. "How many systems/screens does one status answer require today?" 11. "Where are spreadsheets/chat/manual notes still required?"
Capture:
Ask in order: 12. "How do design revisions and change orders hit PM timelines?" 13. "How often do revised specs force schedule or customer-commitment resets?" 14. "How is reprice/reapproval tracked after changes?" 15. "Where is the biggest avoidable loop in that process?"
Capture:
Ask in order: 16. "If we automated one workflow first, what output would you trust on day one?" 17. "What would make you reject an automated status or alert?" 18. "What evidence/metrics would prove this is working in 30 days?" 19. "Who must own adoption on your side for this to stick?"
Capture:
Close script: "Let me play back what I heard as the one highest-value first build. If this is right, we’ll move immediately into a scoped pilot. If not, we correct now."
Ask in order: 20. "Can we get the agreed data set by [date/time]?" 21. "Can we schedule a 30-minute validation review once we map the workflow?" 22. "Any blocker we should remove this week?"
Lock before end:
Go if all conditions are true:
No-go if any condition is false:
Date: 2026-03-02
Prepared by: Patrick Smith / Candlefish.ai
Status: Discovery synthesis (post stakeholder interview)
Date: 2026-03-02
Audience: Executive leadership
Headline: We are paying a recurring tax on manual estimate transfer.
So what: This creates avoidable labor, error risk, and scaling friction.
Speaker notes (30-45s):
The discovery calls show the pain point is not cabinetry expertise. The
pain is that one set of structured data is entered twice across two
systems, with designers acting as the bridge.
Headline: Complexity multiplies cost and quote risk.
Directional annual impact: meaningful recurring labor cost before counting rework/margin leak.
Speaker notes (45s):
Even with conservative assumptions, this is a persistent operational
drag. Bigger jobs drive both higher time cost and higher probability of
quote discrepancies.
Headline: Add an automation layer at the designer estimating stage.
Expected outcome: lower manual entry, lower error surface, better throughput.
Speaker notes (45-60s):
This is a pragmatic integration strategy, not a platform rewrite. We
automate repetitive transfer while preserving expert judgment where
custom work requires it.
Headline: De-risk with phased rollout and measurable gates.
Governance requirements:
Speaker notes (45s):
We sequence this to avoid operational disruption: prove mapping quality
first, then pilot safely, then expand based on results.
Headline: Approve scoped execution now.
Decision needed today: proceed with technical spike and ownership assignment.
Speaker notes (30-45s):
This keeps spend and risk bounded while quickly determining whether we
can materially improve cycle time and quote reliability.
Patrick,
Quick follow-up with a cleaner format for mobile reading.
──────────────────────────────── THARP PACKAGE (READY FOR TODAY) ────────────────────────────────
Attached:
Why this package matters:
──────────────────────────────── 60-SECOND EXEC READ ────────────────────────────────
Build recommendation:
Reason:
Next lane:
Condition:
──────────────────────────────── TODAY’S CALL OBJECTIVE (CHRIS) ────────────────────────────────
By call end, lock:
Output expected same day:
──────────────────────────────── If helpful, I can also send a one-slide visual summary version for immediate forwarding.
Copy/paste prompt for another agent.
You are the principal operations strategist for Candlefish.ai.
Your job is to produce a decision-grade plan for how we proceed after tomorrow's THARP project-management discovery call.
Date context:
- Today: March 4, 2026
- This prompt is for immediate planning around today's THARP PM session.
Known context from prior calls:
[Call A: Design/Pricing SME - Mar 2, 2026, 65 min]
- Participants: Jeanette, Patrick Smith, Randy Long, Tyler Robinson
- Core finding: Cabinet Vision -> Sightline/WIAT handoff at designer stage is manual and error-prone.
- Manual burden: ~5-10 min simple jobs, up to ~1 hour for large multi-room jobs.
- Team volume: ~20-30 jobs/week across ~6-7 designers.
- Common errors: wrong construction/finish transitions and missed overrides.
- Governance: pricing logic maintained in CV material manager; system engineering updates roughly quarterly.
- Data point: MDF export exists; import capability exists in environment but not implemented at the designer step.
- Action signal: THARP agreed to provide exports, work instructions, and sample files.
[Call B: Production Scheduling SME - Mar 3, 2026, 55 min]
- Participants: James Puttmann, Patrick Smith, Ray Khatir, Sam, Tyler Robinson
- Core finding: scheduling truth is partially outside ERP due to unreliable work center dates.
- Current workaround: daily manually rebuilt Excel backlog from ~05:00 export.
- Time burn: Sarah spends ~2-3 hours/day maintaining backlog.
- Risk: single-person dependency (Sarah/Sam); stale daily snapshot; priority based on warehouse date.
- Action signal: backlog files, 15-day build-cycle file, historical exports, and process recording expected.
[Call C: Internal Weekly - Mar 3, 2026]
- THARP confirmed as highest-priority sales/use-case effort.
- Next leadership objective: ROI-focused use case presentation to Santiago and Garth next week.
Your assignment:
Produce a deep, non-generic operational plan for what to do immediately after today's PM session, with explicit decision logic for what to build first.
Hard requirements:
1) Use only evidence-backed claims from the context above.
2) Quantify whenever possible. If data is missing, state exactly what is missing and why it matters.
3) Separate facts, assumptions, and hypotheses.
4) Output must be executable by a mixed team (product, engineering, ops, sales) within 48 hours.
Output format (exact sections):
1. Executive Decision
- One paragraph: "build-first recommendation" and why.
2. Evidence Matrix
- Table with columns:
- Finding
- Source (Call A/B/C)
- Operational impact
- Confidence (High/Med/Low)
- Data gap remaining
3. End-to-End Failure Map
- Step-by-step flow from estimate creation -> drafting -> purchasing -> scheduling -> floor -> customer comms.
- Mark every "human bridge" where data is manually transferred.
- Mark where errors are introduced vs where they are detected.
4. Opportunity Scoring
- Score each candidate from 1-5 on:
- Pain severity
- Universality across custom cabinet shops
- 30-day buildability
- Adoption friction
- Measurability of ROI
- Candidates to score:
- Designer-stage CV -> WIAT estimate sync + validation
- Scheduling backlog autopilot + live exception dashboard
- Quote-to-actual drift/reprice monitor
- PM status automation and customer update layer
5. Tomorrow PM Call Question Set (Priority Ordered)
- 20 questions max.
- For each question include:
- Why this question matters
- Decision it unlocks
- Metric to capture
6. Post-Call Go/No-Go Framework
- Define explicit pass/fail gates for moving into a 30-day pilot.
- Include minimum data/access prerequisites.
7. 30-Day Pilot Blueprint (If Go)
- Week-by-week plan
- Owner roles
- Deliverables
- Risk controls
- Success metrics (baseline -> target)
8. 48-Hour Action Plan
- Numbered tasks with:
- Owner
- Deadline
- Output artifact
9. Risks and Mitigations
- Include technical, operational, adoption, and commercial risk categories.
10. One-Page Brief for Santiago/Garth
- A concise narrative they can read in under 3 minutes:
- Problem
- Why now
- First build
- Expected ROI band
- Ask/next step
Quality bar:
- No generic consulting language.
- No "AI for everything" claims.
- Make tradeoffs explicit.
- Be opinionated when evidence supports a direction.Sessions analyzed: 4 (2 SME interviews + 2 internal debriefs) Source: 1,043 lines of verified transcript across 4 sessions Date range: Mon Mar 2 – Tue Mar 3, 2026 Author: Wake (automated synthesis from transcripts + coaching notes)
Across two completely different departments — Design (Jeanette Carver) and Production Scheduling (Sam Stamp) — we heard the same story told in nearly identical language:
| Session 1: Design/Pricing | Session 2: Production Scheduling | |
|---|---|---|
| Person | Jeanette Carver (CV Lead, 20 yrs) | Sam Stamp (Planning & Scheduling) |
| Broken promise | Sightline import from CV — "never came to fruition" | Work center auto-dates — "never came to fruition" |
| Shadow system | "room block" reference doc + tribal knowledge | "The Backlog" — Excel spreadsheet |
| Manual bridge | CV → Sightline BOM transfer (5-10 min simple, 1hr complex) | Sightline export → Excel transformation (hours/day, ~20 hrs/wk) |
| Bus factor | Jeanette + work instructions (no formal review) | Sam + Sarah only (if both out = "complete halt") |
| Data source | CV Material Manager + designer memory | "Job Orders Master Export" from Sightline |
| Exact quote | "We were promised that — it just never came to fruition" | "We were promised that and never came to fruition" |
The pattern: SyteLine/Wyatt (Infor ERP, also called N4) consistently promises automation features that are never properly implemented. Each department independently builds manual workarounds that become mission-critical, fragile, and person-dependent.
Built entirely from transcript statements (not inferred):
DESIGN PHASE
┌─────────────────┐ MANUAL BRIDGE ┌─────────────────────┐
│ Cabinet Vision │ ──────────────────────→│ Sightline / Wyatt │
│ (CAD/Design) │ 5-6 designers │ (Infor ERP) │
│ │ 20-30 jobs/week │ │
│ Material Mgr │ 5-10 min simple │ Estimates │
│ = pricing truth │ up to 1hr complex │ Opportunities │
└─────────────────┘ "room by room" └────────┬────────────┘
│
Customer approval
│
▼
┌─────────────────┐
│ Drafting │
│ ("One Unit" / │
│ "Whistles") │
│ BOM file runs │
│ here │
└────────┬────────┘
│
Job loaded into Wyatt
Status → "Firm"
│
PRODUCTION PHASE ▼
┌─────────────────┐ 5 AM EXPORT ┌─────────────────────┐
│ "The Backlog" │ ←──────────────────────│ Sightline Workbench │
│ (Excel) │ "Job Orders Master │ (Firm → Released) │
│ │ Export" + hours of │ │
│ REAL source │ manual manipulation │ Work center dates │
│ of truth │ by Sarah │ = "all completely │
│ │ │ wrong" │
│ $1.62M WIP │ └────────┬────────────┘
│ tracked here │ │
└────────┬────────┘ Live sync
│ │
│ Afternoon meeting ▼
│ w/ dept leads ┌─────────────────┐
│ (manual patch notes) │ Factory Track │
│ │ (iPads on floor)│
▼ │ Dept sign-outs │
┌──────────────┐ └─────────────────┘
│ 15-Day Belt │
│ Cycle │
│ │
│ Days 1-5: │
│ Panel │
│ Doors │
│ Rough Mill │
│ │
│ Days 5-7: │
│ Specials │
│ │
│ Days 7-15: │
│ Finish │
│ Assembly │
│ │
│ → Warehouse │
│ → Delivery │
└──────────────┘Key insight: There are TWO BOM paths. Jeanette's manual transfer is for pricing/estimating. Drafting creates the production BOM when loading into Wyatt. These are separate — pricing errors don't directly cause production floor errors, but they do cause customer-facing quote errors.
All numbers sourced directly from transcripts:
| Metric | Value | Source |
|---|---|---|
| Designers doing manual transfer | 5-6 | Jeanette |
| Jobs per week | 20-30 | Jeanette |
| Time per simple job transfer | 5-10 minutes | Jeanette |
| Time per complex job transfer | Up to 1 hour | Jeanette |
| CV performance degrades after | 8-10 rooms per job | Jeanette |
| Ship-to addresses per contractor | Up to 40-50 | Jeanette |
Estimated weekly labor (conservative): 5-6 designers × 25 jobs × 10 min avg = 25 hours/week on manual transfer alone
| Metric | Value | Source |
|---|---|---|
| Backlog rebuild time (Sarah) | ~50% of her week (~20 hrs/wk) | Sam |
| WIP on production floor | $1.62 million | Sam (from Backlog screen share) |
| Weekly floor release target | $500K | Sam |
| Sample week: released vs. completed | $350K released / $444K completed | Sam |
| Belt cycle | 15 days | Sam |
| Warehouse date buffer (normal) | 5 days before delivery | Sam |
| Warehouse date buffer (busy) | 2 days before delivery | Sam |
| People who can build the Backlog | 2 (Sam + Sarah) | Sam |
| Backlog refresh frequency | Once at 5 AM, stale all day | Sam |
| Work center dates accuracy | 0% — "all completely wrong" | Sam |
| Metric | Value | Source |
|---|---|---|
| Automatable hours (2 divisions only) | 50-60/week | Patrick (back-of-envelope) |
| Estimated annual labor cost | "Well north of a million" | Patrick |
| Potential labor cost (full company) | "$2 million" | Patrick |
| Tharp estimated revenue | ~$50 million | Patrick (from financials seen in Backlog) |
Evidence strength: ★★★★★ (confirmed by 2 SMEs + Sightline import exists + universality confirmed)
Evidence strength: ★★★★★ (screen-shared, process documented, time quantified)
Evidence strength: ★★★★☆ (confirmed broken, root cause unknown — Kris Axtell session may clarify)
Evidence strength: ★★★☆☆ (need confirmed, mechanism unclear)
Evidence strength: ★★★☆☆ (error risk confirmed, frequency unquantified)
| Signal | Source | Strength |
|---|---|---|
| Jeanette confirmed same problem at prior CV shop | Session 1 transcript | ★★★★★ |
| "Reality of doing custom cabinetry" | Jeanette, Session 1 | ★★★★☆ |
| Sam: smaller shops don't have ERPs, fully manual | Session 2 transcript | ★★★☆☆ |
| Garth "not a tech-forward operator" — if Tharp operates this way, peers will too | Ray, Debrief 2 | ★★★☆☆ |
| "There's no way every cabinet maker isn't encountering these" | Tyler, Debrief 2 | ★★★☆☆ |
| "Call 50 and 20 of them have the same problem" | Ray, Debrief 2 | ★★☆☆☆ (hypothesis, not validated) |
| CV is "the thing" — most custom cabinet shops use it | Jeanette, Session 1 | ★★★★☆ |
Assessment: Strong signal that the CV→ERP gap is universal among growth-stage custom cabinet shops using Cabinet Vision. The production scheduling gap (Backlog/work center dates) is less clearly universal — may be Sightline/Infor-specific. Two more sessions (Chris Sober, Kris Axtell) will clarify.
Not yet validated: The TAM claim of "~5,900 US cabinet shops" appears in earlier docs but has no source in any transcript. It should be treated as an assumption until independently verified.
| Option | Description | Source Quote |
|---|---|---|
| Custom build for Tharp | 4-5 workflow automations, $50-100K each | "Four or five workflows and they're willing to pay 50,000 or 100,000" — Patrick |
| Subscription model | $8K/month vs. hiring $200K FTE engineer | "Better for you to pay us $8,000 a month for the software" — Patrick |
| Phase 1 → Phase 2 | Build for Tharp (immediate ARR) → productize → cold call other shops | "Build immediate ARR... so you can focus on Phase 2" — Patrick |
| Free first, upsell rest | First integration free to prove it works, price subsequent ones | "Start with one component for free just to prove it out" — Patrick |
| Cold call validation | Max does 100 cold calls to cabinet shops | "Call 50 and 20 of them have the same problem" — Ray |
Current consensus (Debrief 2): Don't build yet. Finish all sessions. Then quantify total automatable labor cost across all departments, present to Santiago with specific proposals.
This document synthesizes only verified information from transcripts and coaching notes. Claims are attributed to specific people and sessions. Unverified assumptions are flagged. Updated after each session.
Both Session 1 and Session 2 were captured by Read.ai with speaker-attributed transcripts, AI summaries, and action items. On Mar 3, all uncertain terms and claims in this document were cross-referenced against the Read.ai transcripts.
| Our Term | Read.ai Actual | Source |
|---|---|---|
| "Shackle Groups" | rate tables (6 different) | Jeanette, Session 1 |
| "Room Lock" | room block (CV reference screen) | Jeanette, Session 1 |
| "NDF export" | CSV/Excel export ("export bid data") | Jeanette + Randy, Session 1 |
| "cab cam report" | cab count report | Jeanette, Session 1 |
| "45-page training doc" | work instructions (written 8 yrs ago, no page count) | Jeanette, Session 1 |
| "One Unit" | Not a real term — Whisper artifact. Sam says "Drafting" | Sam, Session 2 |
| Claim | Read.ai Confirms | Source |
|---|---|---|
| 5-6 designers | ✅ "five or six designers" | Jeanette |
| 5-10 min simple transfer | ✅ "five or 10 minutes" | Jeanette |
| Import promised, never implemented | ✅ "phase two of sightline implementation" | Jeanette |
| "We were promised that" | ✅ Verbatim in both sessions | Jeanette + Sam |
| Ray absent from Session 1 | ✅ attended=False in Read.ai | Participant data |
| James + Ray present Session 2 | ✅ attended=True | Participant data |
| Sarah rebuilds at 5 AM | ✅ "05:00 Sightline pull" | Read.ai summary |
| Work center dates broken | ✅ "routinely incorrect" | Read.ai summary |
| Randy says CSV export possible | ✅ "CSV and it should just be mapped" | Randy, Session 1 |
| Claim | Status |
|---|---|
| $1.62M WIP | Not in Read.ai text; was during screen share (our Whisper captured it, Read.ai may not have) |
| 5,900 TAM | No source in any transcript or email |
| Kris Axtell's exact role | Inferred from session topic |
Status: In progress — 2 of 4 SME sessions complete (+ 2 internal debriefs) Last updated: Tue Mar 3, 2026
Custom cabinet shops at growth stage run on ERPs that consistently under-deliver. The result: every department builds manual workarounds that become mission-critical single points of failure. The pattern is identical across design (CV→Sightline bridge), production scheduling (the Backlog), and work center tracking (broken auto-dates). This is not a Tharp problem — it's a growth-stage cabinet manufacturing problem.
Full summary:
meetings/2026-03-02-tharp-cv-pricing/summary.md
Core finding: 5-6 designers manually transfer BOM data from Cabinet Vision → Sightline/Wyatt for every job. 20-30 jobs/week. Simple job = 5-10 min, complex = up to 1 hr. No validation gate before customer sees quote. Wrong construction type = drastic price difference.
Key intel:
Universality: Jeanette confirmed from prior employer — same problem at another CV shop. "Reality of doing custom cabinetry."
Quote: "We were promised that — it just never came to fruition."
Full summary:
meetings/2026-03-02-tharp-debrief-tyler/summary.md
Strategic alignment:
Full summary:
meetings/2026-03-03-tharp-production/summary.md
Core finding: Production scheduling runs on a manually-rebuilt Excel spreadsheet ("the Backlog") because Sightline's work center dates never worked. Sarah rebuilds it every morning at 5 AM from a raw Sightline export — takes ~20 hrs/week. $1.62M WIP tracked manually. Stale by noon. If both Sam and Sarah are out, production has no visibility.
Key intel:
Universality: Partial. Smaller shops are fully manual (no ERP). Tharp is in the growth trap — big enough to buy ERP, not big enough to get it properly implemented. Market = growth-stage shops with ERP + manual workarounds.
Full summary:
meetings/2026-03-03-tharp-debrief-internal/summary.md
Key strategic points:
| Date | Time | Topic | SME | Status |
|---|---|---|---|---|
| Wed Mar 4 | 11:00 AM | PM / Workflow | Chris Sober | 📋 Prepped |
| Wed Mar 4 | 2:00 PM | CV→Wyatt Integration / DRI | Kris Axtell | 📋 Prepped |
| Name | Role | Notes | |
|---|---|---|---|
| Santiago Ramirez | COO (champion) | sramirez@tharpcabinets.com | Cell: (970) 698-0377 |
| Jeanette Carver | Designer/CV Lead | jcarver@tharpcabinets.com | 20 yrs CV experience, trains all designers |
| Randy Long | Pricing/Materials | rlong@tharpcabinets.com | Manufacturing background, wants structure |
| Sam Stamp | Production Scheduling | sstamp@tharpcabinets.com | Runs 15-day belt cycle w/ Sarah. Owns the Backlog |
| Chris Sober | Project Management | csober@tharpcabinets.com | Wed Mar 4 11AM session |
| Kris Axtell | CV→Wyatt Integration / DRI | kaxtell@tharpcabinets.com | Wed Mar 4 2PM (role inferred from session topic — confirm) |
| Garth Rummery | Leadership | GRummery@tharpcabinets.com | Executive sponsor. Not in SME sessions. |
| Session 1 (Jeanette/Randy) | Session 2 (Sam) | |
|---|---|---|
| Broken promise | CV→Sightline import | Work center auto-dates |
| Shadow system | room block doc | The Backlog (Excel) |
| Manual bridge | BOM transfer (5-10 min/job) | Daily 5AM rebuild (~20 hrs/wk) |
| Bus factor | Jeanette (training doc = tribal knowledge) | Sam + Sarah (only 2 can build Backlog) |
| Quote | "We were promised that — it just never came to fruition" | "We were promised that and never came to fruition" |
| Universality | Confirmed (prior employer) | Growth trap (smaller shops fully manual) |
This document is internal to Candlefish. Updated after each session. Will become the basis for the final recommendation to Garth Rummery after Wed Mar 4.
Wake is listening. Questions ranked by automation potential for Candlefish. Glance here when there's a natural pause in conversation.
[14:00] 🟢 Call starting — listening for pain points
Openers to keep in your back pocket if conversation stalls: → "Before we dive into systems — can you each walk me through what a typical day looks like for you?" → "When a new kitchen remodel inquiry comes in, who touches it first and where does that info go?"
[14:08] 🔥 "Infor over promised" — SME expected system capability that doesn't exist
→ "What specifically were you told Infor could do that it can't? Was that at purchase or during implementation?" → "So how are you handling that gap right now — manual process, workaround, or just living with it?" → "How much time per week would you estimate that workaround costs you?"
💡 This is exactly the gap-between-systems play. Quantify the time cost.
[14:10] 🔥🔥🔥 Jeanette manually transfers BOM from Cabinet Vision → "Sightline" (not N4?). 5-10 min simple job, longer for complex.
→ "Sightline — is that the same as N4/Infor or a different system? How does it connect to the ERP?" → "How many jobs per week are you doing this transfer for?" → "When you say 'manually transfer' — are you retyping from a screen, copy-pasting from an export, or something else?" → "What are the most common errors that creep in during that manual transfer?" → "For a complex job with different woods/finishes per room — how long does THAT take?"
💡 THIS IS THE #1 BUILD TARGET. CV→Sightline manual BOM transfer. Quantify: jobs/week × time/job = hours saved. Error rate = quality win. Every cabinet shop with CV has this pain.
[14:11] 🔥 Jeanette screen-sharing CV. Reports tab → Summary → must manually select "rate tables" + construction type + finish level. Multiple pricing tables with different levels.
→ "Those construction and finish selections — is that always the same set of choices, or does it change? Could a system auto-detect which one to use based on the design?" → "How many different pricing tables / construction levels are there total?" → "What happens if someone picks the wrong construction or finish level — does the pricing just come out wrong silently?"
💡 Manual selection from dropdowns before pricing runs = automatable decision tree. If choices are deterministic from the design, this is a rules engine play.
[14:13] 🔥 Each room priced SEPARATELY in CV. 20 rooms = bounce back and forth selecting construction + finish for each one individually. "Finicky" system.
→ "So for a 20-room job — you're running this pricing report 20 separate times, once per room, each with its own selections?" → "About how long does a 20-room job take you start to finish just for the pricing step?" → "Has anyone ever accidentally priced a room with the wrong finish level and it went out to the customer that way?"
💡 MASSIVE. N rooms × manual selections × error risk = combinatorial pain. A tool that reads the design specs and auto-selects construction + finish per room eliminates this entirely. Universal across all CV shops.
[14:14] 🔥🔥 KEY: "Sightline" = what they call "Wyatt". She's now switching to the estimates form in Sightline/Wyatt. CV slows down significantly after 8-10 rooms. Multiple finishes in ONE room = must break out and price separately.
→ "So Sightline and Wyatt are the same thing? And that's separate from N4/Infor?" → "When you move from CV to Sightline — are you manually retyping what CV just showed you, or is there any export/import?" → "You mentioned CV slows down after 8-10 rooms — is that a common complaint across your team?"
💡 NOW WATCHING THE HANDOFF. She's about to show how data moves from CV → Sightline/Wyatt. This is the exact integration gap. Watch for: manual retyping, copy-paste, or file export.
[14:15] Now in Infor/Sightline. Salespeople create the opportunity first. Jeanette creates estimate attached to it. Must pick from ship-to addresses — contractors can have 40-50 addresses.
→ "So the salesperson sets up the opportunity in Infor, then you come in and attach the estimate — is there anything from their setup that auto-populates into your estimate, or are you starting fresh?" → "Those 40-50 ship-to addresses for contractors — is that a pain point? Do people pick the wrong one?" → "When you're building this estimate, are you looking at CV on another screen and typing these numbers in?"
💡 Two-system workflow confirmed: sales creates opportunity in Infor, designer creates estimate, then manually bridges CV pricing into it. Watch for the moment she starts entering the actual pricing data.
[14:16] 📋 FULL WORKFLOW SO FAR — use this to play back to Jeanette:
"Let me make sure I've got this right so far:
Is that right? Am I missing anything?"
[14:17] 📋 UPDATED WORKFLOW — new steps added:
Step 7 (refined): In Infor, estimate gets an ESTIMATE NUMBER generated Step 8: Go into ESTIMATE LINES — create a separate line for EACH ROOM Step 9: Pick room type from a list (kitchen, bathroom, etc.) Step 10: Name each room in description field ("Bathroom 1", "Bathroom 2", etc.) Step 11: (coming next) — manually enter the pricing/materials from CV for each line
→ "So each room becomes its own line item in the estimate — and you're typing in the room names to match what you designed in CV? Does CV and Infor share any room naming or numbering, or are you matching them up by memory?"
💡 No shared identifier between CV rooms and Infor estimate lines = manual matching by name/memory. That's an error vector AND the exact thing an integration bridge would solve.
[14:18] 📋 WORKFLOW UPDATE:
Step 10 (refined): She has a "ROOM LOCK" — a reference doc she uses to remember what to enter for each room Step 11: Now manually filling out fields for each estimate line — picking door style, finish, construction from dropdowns in Infor Step 12: (in progress) Selecting through multiple fields per room — door, finish, etc.
→ "That room block reference — is that something you printed from CV, or is it notes you took separately?" → "How many fields do you fill out per room in this estimate? Ballpark — 5? 10? 20?"
💡 She needs a CHEAT SHEET just to remember what to type from one system to the other. That's the definition of a manual bridge. Count the fields per room — that's your automation scope.
[14:19] 📋 COMPLETE WORKFLOW — Jeanette's pricing journey so far:
SYSTEM 1: Cabinet Vision (CAD/Design)
THE GAP (manual bridge) 10. Jeanette looks at CV output (room block/reference) and carries data in her head or notes to the next system
SYSTEM 2: Sightline/Wyatt (= Infor ERP) 11. Find the opportunity (created by salesperson) 12. Create new estimate → gets estimate number 13. Pick ship-to address (contractors can have 40-50) 14. Create ESTIMATE LINES — one per room 15. Pick room type from dropdown, name it in description 16. Fill out fields per room: door style, construction, finish, material 17. Save each line 18. Repeat for every room in the job
PAIN POINTS IDENTIFIED:
[14:23] ✅ She confirmed the workflow. "Yeah" to being the manual bridge.
NOW PIVOT TO QUANTIFYING + ERRORS + WHO ELSE:
→ "How many jobs per week are you doing this for? And is it just you or does the whole team do this same process?" → "When something goes wrong — wrong finish, wrong construction — how far does that error get before someone catches it? Does it make it to the floor? To the customer?"
THEN THE MAGIC QUESTION: → "If we could make it so that when you finalize in CV, all those room details just appeared in Sightline automatically — no retyping, no room block — what would that change for you?"
💡 Let her answer that last one emotionally. Her words = your pitch to Garth.
[14:25] 🔥🔥🔥 FIVE OR SIX DESIGNERS all doing this same manual bridge. Plus a new hire starting today.
KEY DATA CAPTURED:
→ Now she's about to answer the error question. LISTEN for:
💡 5-5-6 designers × multiple jobs/week × manual transfer = MASSIVE total hours. This is the scale story for Garth AND for the 5,900 shops nationally.
[14:26] 🔥 Wrong construction type = DRASTIC price difference. Errors CAN reach the customer quote. Some salespeople worse at catching it.
→ "When the construction type is wrong and the price is way off — has a customer ever caught that and you had to re-quote? What does that look like?" → "You said some salespeople aren't as good at catching errors — is there any validation step between your estimate and what goes to the customer, or does it go straight out?" → "Randy — from a pricing perspective, how big is the price swing between construction types? Like percentage-wise?"
💡 Construction type error = silent pricing bomb. No validation layer between Jeanette's manual entry and customer-facing quote. This is the error-checking automation play — a rules engine that validates CV selections against Sightline entries BEFORE the quote goes out.
[14:27] 📋 ASK TO MAP OWNERSHIP:
→ "Can you help me understand who owns each piece of this? Like:
💡 Map the handoff chain: Sales → Design → Pricing → Review → Customer. Every handoff without validation = error risk. And getting Randy to define his role clarifies whether pricing ownership is centralized or distributed across 5-6 designers.
[14:28] 🔥 After designer finishes + customer approves → goes to DRAFTING team. Drafting finalizes, creates CNC programs, uploads to shop floor. Designer "should" double-check but it's on them — no formal review gate.
→ "So the designer is essentially self-checking their own work? There's no second set of eyes before it goes to the customer?" → "Drafting — when they finalize and create the CNC programs, are they pulling from CV or from Sightline? Or both?" → "Randy — is the pricing locked at that point, or can it still change after drafting gets it?"
💡 NEW SYSTEM in the chain: Drafting uploads to shop floor (CNC). That's CV → Sightline → Drafting → CNC. Three handoffs. No formal review gate — designer self-checks. Drafting is a third persona to potentially interview.
[14:29] 🔥🔥🔥 "We were promised that — it just never came to fruition."
QUOTE OF THE MEETING. They were TOLD this integration would exist. Infor promised it. It never shipped. This is the Unburdn story repeating — promises without delivery. Candlefish's edge is EXECUTION.
NOW PIVOT TO RANDY & THE PRICING SIDE: → "Randy — I want to make sure we hear from you too. On the pricing side, where do the pricing tables and rate structures live? Are those in CV, in Sightline, or somewhere else?" → "When material costs change — lumber goes up, hardware supplier changes pricing — how does that flow into what the designers are quoting?" → "How often do quoted prices end up different from actual production costs?"
💡 You've mapped Jeanette's pain deeply. Now shift to Randy before time runs out. Pricing data flow + cost variance = second automation surface. Don't leave without Randy's perspective.
[14:30] Confirmed: no second set of eyes. Designer self-reviews.
REMAINING JEANETTE QUESTIONS (pick 2-3 before pivoting to Randy):
→ "When a customer wants changes mid-process — say they switch from oak to maple on two rooms — how much rework does that create? Do you redo the whole thing in CV AND Sightline?" → "How many revisions does a typical job go through before the customer signs off?" → "After a job ships, do you ever compare what was quoted vs what it actually cost to produce? Or does that happen somewhere else?" → "If you could wave a magic wand and fix ONE thing about this whole process, what would it be?"
💡 The revision/change-order question is critical — if every customer change means double rework (CV + Sightline), the time multiplier is even bigger than the initial entry.
[14:32] 🔥 Randy says: CV already HAS the data — "it's just not a first step." There IS an existing process further along that does some transfer, but it's LATER in the workflow (after design, in the drafting/production phase).
→ "Randy — so there IS a connection that works later in the process? What does that look like? What data does it move and between which systems?" → "Why does that connection work at the production stage but not at the estimate stage? Is it a different export format, different data, or just how it was set up?" → "If that later-stage transfer already exists, could the same mechanism be pulled forward to the estimating step?"
💡 CRITICAL INSIGHT. An integration path ALREADY EXISTS but only for production, not for estimating. The data format and pipeline may already be partially built. This dramatically reduces the build scope — you may not need to build from scratch, just extend what's there.
[14:34] Randy clarified the handoff: Designer → Salesperson confirms price → "ships off" to Drafting → Drafting uses "Whistles" to turn the design into full blueprints. All still in CV ("CME"). Jeanette's design work IS the drafting team's starting point — they take it to "the next level."
Customer change (oak→maple): Jeanette says "oh no" — sounds like it's NOT a full redo. Let her explain.
→ "So the change itself in CV is quick — but do you then have to go back into Sightline and update the estimate too?" → "How many revision cycles does a typical job go through before the customer signs off?"
💡 If CV changes are easy but the Sightline re-entry is the painful part of revisions, that strengthens the bridge automation case even further — every revision = another manual transfer.
[14:36] Templates exist in CV but they're inconsistent — Tharp creates some, designers create their own custom ones. Even after selecting a template, Jeanette still has to manually change certain fields. "Global issue" — some settings override at the job level but others don't, requiring manual fixes. Tribal knowledge: "you know through experience, if you don't know..."
→ "So even with templates, there's tribal knowledge about which fields you have to manually override? Is that written down anywhere, or does each designer just learn it over time?" → "For the new designer who started today — how long does it typically take before someone is fully up to speed on all these workarounds?" → "How many of these manual overrides would you estimate there are per job? Like 5? 20?"
💡 Templates EXIST but don't fully work — manual overrides required through tribal knowledge. New designer ramp time = training cost. This is a configuration/rules engine play on TOP of the bridge. Codify the tribal knowledge into validation rules.
[14:37] 🎯 UNIVERSALITY PROBE — keep it casual, don't tip that you're sizing a market:
→ "Do you guys ever talk to other cabinet shops — like at trade shows or Cabinet Vision user groups? Do they deal with this same stuff?" → "When you went through CV training, was this manual transfer to the ERP something they addressed, or was it just kind of assumed you'd figure it out?" → "Is this a Cabinet Vision thing specifically, or do you think shops using other CAD tools have the same gap?"
SOFTER ALTERNATIVES: → "Do you think this is a Tharp-specific challenge because of your scale, or is this just the reality of custom cabinetry?" → "Has Hexagon (the CV maker) ever talked about building this integration, or is it just not on their radar?"
💡 Frame it as curiosity about the INDUSTRY, not about market sizing. Her answer tells you if this is Tharp-specific or universal. If she says "everyone deals with this" — that's your 5,900-shop TAM confirmed from the practitioner's mouth.
[14:39] 🔥🔥🔥 UNIVERSALITY CONFIRMED. Jeanette worked at ANOTHER custom cabinet shop that also used CV. Same problem. "Reality of doing custom cabinetry." CV is "the thing" — most people using it.
→ "At that other shop, how did they handle the CV-to-ERP gap? Same manual process, or did they have a different workaround?" → "Randy — from your side, do you hear about this from other people in the industry?"
💡 JACKPOT. She just confirmed from first-hand experience at a different shop: same tool (CV), same gap. This is NOT Tharp-specific. It's an industry-wide problem. You now have practitioner-level TAM validation.
[14:40] 🔥 Previous shop: exported CV report to EXCEL, then manually double-checked sizes/percentages. Same manual bridge, different flavor. Distributors using basic drafting software don't even have CV.
KEY INTEL:
→ "That Excel export from CV — does Tharp use that too, or do you skip that step?" → "What format does CV export to? Is it always Excel, or can it do CSV, XML?" → "Randy — on the pricing side, are you maintaining pricing tables in Excel too, or is that all in Sightline?"
💡 CV exports to Excel = there IS a structured data output. If we can read that export format, we can build the bridge without needing CV's internal API. The export file IS the integration surface. Find out exactly what's in it.
[14:41] Randy talking about pricing MODEL complexity. CV handles design + layout, they're trying to put pricing within it but "a little bit too complex." Sounds like pricing logic doesn't fit neatly inside CV.
→ "Randy — where does the actual pricing logic live today? Is it built into CV's rate tables, or do you maintain it separately?" → "When you say the pricing model is too complex for CV — what makes it complex? Is it the number of variables, the finish combinations, or something else?" → "If the pricing rules lived outside of CV — like in a separate engine that could read the design and calculate the price automatically — would that be better or worse for you?"
💡 Randy's pain is different from Jeanette's. She's doing manual data entry. He's dealing with pricing LOGIC that's too complex for the tool. A pricing rules engine that sits BETWEEN CV and Sightline = solves both problems simultaneously.
[14:42] Randy's background is manufacturing/production — "advanced action, experience and repeat." He came from a world of STANDARD sizes. Custom cabinetry = every cabinet is different dimensions. He's trying to impose structure/repeatability onto a non-standard process. "Slow down, double check everything."
→ "Randy — it sounds like you came from a more standardized manufacturing environment. What's the biggest thing that breaks when you try to apply those same principles to custom work?" → "When you say slow down and double check — what are you checking against? Is there a master spec, or is the CV design the source of truth?" → "Between you and Jeanette — who's the final authority on whether the pricing is right before it goes to the customer?"
💡 Randy is a process/systems thinker in a chaos environment. He WANTS structure but the tools won't give it to him. A validation engine that catches errors before they propagate = speaks directly to his DNA. Frame the build as "giving you the double-check automatically."
[14:44] 📋 RANDY'S PRICING WORLD — QUALITATIVE SUMMARY:
Where pricing lives:
The chain: Randy (identifies price changes) → Management (approves) → Systems Engineers (update Material Manager in CV) → Designers (use updated prices when they run reports)
What this means:
→ "How often do prices get updated in Material Manager? Monthly? Quarterly? When suppliers change?" → "Between when Randy flags a price change and when it actually shows up in CV — how long is that lag?" → "Has a quote ever gone out based on old pricing because the update hadn't been made yet?"
💡 The lag between price change identification and CV update = quotes going out on stale data. That's a pricing accuracy problem every shop has.
[14:45] 📋 FULL STRATEGIC PICTURE — WHAT WE'VE HEARD:
#1 — The CV→Sightline Bridge (HIGHEST VALUE)
#2 — Pricing Validation Engine
#3 — Price Update Propagation
[14:46] 🎯 MUST-ASK QUESTIONS BEFORE YOU WRAP:
Technical (need for the build): → "Can we get a sample CV export file — the Excel or report output? Just a dummy job, nothing confidential. That tells us exactly what data comes out and what format." → "Does Sightline/Wyatt have an API or any import capability — CSV upload, anything?" → "Who are the Systems Engineers we should talk to about CV's data structure?"
Quantitative (need for the business case): → "Roughly how many jobs per week across all 5-6 designers?" → "What's the average number of rooms per job?" → "When an error makes it to the customer and you have to re-quote — how often does that happen? Monthly? Weekly?"
Relationship (need for continued access): → "Would you two be willing to test a prototype if we came back with something in a few weeks?" → "Can we do a screen-share session where we watch you do one real job start to finish?" → "Is there anyone else at Tharp — like the Systems Engineers or Drafting — we should talk to?"
💡 The sample export file is the SINGLE most important ask. That's your spec for the build. Don't leave without it or a commitment to send it.
[14:49] 🔥🔥🔥 KEY DATA CAPTURED:
MATH:
→ "That import capability in Sightline — do you know what format it accepts? CSV? XML?" → "Would the Systems Engineers be the right people for us to talk to about getting that import working?" → "Can we get a sample of that CV export and a sample of what Sightline expects on import?"
💡 GAME CHANGER: Import exists but isn't set up. This could be a configuration + mapping project, not a from-scratch build. Dramatically reduces scope. GET THE FORMAT SPECS.
[14:50] 📋 CLOSING PLAYBACK — read this to wrap:
"Here's what I'm taking away from today, and I want to make sure you both agree:
Jeanette, your superpower is design quality and accuracy — 20 years in CV, you know what looks good and what the limits are. Very few errors make it past you. But the system is forcing you to spend your time on data entry instead of design. You and five other designers are manually bridging Cabinet Vision and Sightline 20-30 times a week — selecting construction types, finishes, creating estimate lines room by room — when that data already exists in CV.
Randy, you're bringing manufacturing discipline to a custom world. You want structure, validation, repeatability — but the tools make you slow down and double-check everything manually because there's no automated guardrail.
The good news: Sightline already has import capability that just hasn't been set up. CV already exports structured data. The bridge between these systems isn't a technology problem — it's an implementation problem. And that's exactly what we build.
Here's what we'd like to do next: get a sample CV export and the Sightline import spec from your Systems Engineers, and come back to you in a few weeks with a working prototype that takes a CV job and pushes it into Sightline automatically — rooms, pricing, materials, all of it. No retyping. No room block. Would you be open to testing that?"
[14:52] 📋 REMAINING THARP SESSIONS THIS WEEK:
SAY: "This is one of several sessions we have this week. Tomorrow we're meeting with Sam on Production Scheduling, Wednesday with Chris on Project Management and Workflow, and then a session specifically on the CV-to-Wyatt integration. Everything you've shared today feeds directly into those conversations — especially that integration session on Wednesday."
[14:57] 🏁 CLOSING — you've asked "anything else?" Perfect. After they respond, wrap with:
"This has been incredibly valuable. You've given us exactly what we need to start working on this. A few things from our side:
We'll set up that shared Google Drive folder and send you both an invite — just drop in anything you're comfortable sharing. The CV export, the rules, any reference docs.
We'll take everything from today and map it against what we hear in the other sessions this week.
We'll come back to you — probably in two to three weeks — with a specific recommendation: here's the one thing we'd build first, here's what it does, here's what it saves.
No commitment, no cost to Tharp. We just need your expertise to make sure we build the right thing.
Thank you both — this was exactly the conversation we needed to have."
[13:58:11] Can you show me a sample of the coaching terminal questions that you're going to help me with?
[13:58:33] Shade Shade Shade Shade Shade Shade Shade Shade Shade Shade Shade Shade Shade Shade Shade Shade
[13:58:59] 어요
[14:07:43] I'm like, I thought that was supposed to be part of the system, that we were supposed to be able to do that. I don't know if the informer over promised, but that's my biggest one.
[14:08:01] So what are you doing instead? Like what's the actual workflow?
[14:08:18] for lack of a better word with the standalone build materials. So it's unique every single time. So we might have common material between them, but trying to force it in.
[14:08:35] and custom bill of materials and not a repeatable build process makes that that much more of a challenge. So, to a point like uploading, you know, this person wants it in.
[14:08:52] and bright orange or something like that. Just the certain things, it's not as easy as just saying, here's iPhone 17 and we're gonna build all the green ones the same damn way. No, that makes a ton of sense. Yeah, so.
[14:09:10] Is it just like this manual process this work around or are you just kind of living with it?
[14:09:27] And it also would be helpful for us like if you have a sense, Jeanette, of like the time impact associated with this for you, like how much time per week you estimate that worker.
[14:09:44] Um, you know, for a simple job where, you know, what I call a simple job where even if it's, you know, five or ten rooms, each, even if the whole house is the same.
[14:10:01] It only takes like 5 or 10 minutes to go through and import, manually transfer the data from Cavavision to Sightline. If it's a big job, I mean...
[14:10:19] Each room has different woods and finishes and I just want to try to pull this up and then maybe Yeah, great. Yeah, that'd be awesome. And obviously when I'm...
[14:10:36] something very basic. Okay, great. Yep, let's show it out.
[14:10:53] Yeah, so this is CAG Division. I've just, you know, quickly drawn along, placed the cabinet. What we have to do is we come into this reports tab and then we get this summer.
[14:11:10] And then we have to make sure that everything we want to price is selected in rate tables. And then we have these rig tables where you can, you know, we always...
[14:11:27] always pick that one. But then there are these different instructions that we have and they all have different pricing levels and so you have to select which construction you're in and then you have to pick which finish they're doing.
[14:11:45] when you clicked on. Out. When you clicked on add-ons. Yeah, I'm really like that. Well maybe this is part of it, right? The system in and of itself is a little bit finicky.
[14:12:02] Yeah, oh you see it okay, so this is just screen here. Oh, funny, okay.
[14:12:19] instead of the, yeah, if you're right. Um, I want to show you that screw. Let's do that. And turn off all the other stuff. And then, okay, now you can see it. There we go.
[14:12:37] and you basically have to make a selection for each one based on what the selection is for the rumor. And if you're switching that towards the tweaking of the room,
[14:12:54] or you know you have 20 rooms and they all have different finish levels. You have to bounce back and forth because if you get different rooms, I guess I could show that too if I had a second room.
[14:13:11] Because we prank each other separately. So, you know, like community reports.
[14:13:28] and then fat fat not, if I come into a big center, and then I go into filters, you can see different rooms. And sometimes there will be multiple finishes in one room, which means you have to break
[14:13:45] and I know selects the next cabinets for them. It can get really tedious. Yeah, yeah. And you're not seeing it, but once you get to about eight or 10 runs in a job, Cavavision really,
[14:14:03] you know right now it's working pretty quickly because I've only got two rooms and two cabinets so once I have you know five ones selected i kind of kind of kind of kind of kind of kind of kind of kind of kind kind kind kind kind kind kind kind kind kind kind kind kind kind kind kind kind kind kind kind kind kind
[14:14:22] So this is a site line, what we call Wyatt, and I would come into the estimates form and maybe I should open.
[14:14:39] the training so that I'm not trying to create something that isn't actually going to be used. So let me get into our training of this. Cause I can just run in there all I want to do. This is all super helpful about this is exactly what we're looking for.
[14:14:56] Okay, so I'm just going to go take an opportunity. Estimates have to be attacked.
[14:15:14] the salespeople set up the opportunity so I can't, I still with that but so you just find your opportunity and you create an estimate and then I have to pick the
[14:15:31] If we have contractors who do multiple jobs with us, you know, I've seen with 40 or 50 ship to addresses. So I said that I think.
[14:15:48] And then I... Oh wait, there's already a smoothing here. Just like... No. I would create a new one. And then...
[14:16:05] thank you thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank thank
[14:16:24] Now we should be able to save. And the kernel database does run slower than the production database, so it's not usually...
[14:16:41] But then it creates an estimate number and then I go into estimate lines. And I have to create a line for each, typically each room or each.
[14:16:58] So, for example, all that data. You know, we have different kinds of rooms set up, you know, without building a room.
[14:17:15] So I'll just pick that and then I can name the room here so that you know if I have three bathrooms and I'll just say bathroom bathroom bathroom I can just name it in description.
[14:17:33] It's the master path, double, undefended, and the next thing I want to do is one. And then we have to go into the student context, and we have to use the...
[14:17:50] That is the end of the year that we set out the job with. Which we have half of the year. I wouldn't be wrong about what I wrote.
[14:18:07] So this is our room block. That's just gonna make it easier for me to remember what I need to take. Come back here and then yeah, I have to fill out.
[14:18:24] Hey, today's the next hour. What happened? Oh, I... I just did a... There we go. Sorry. I should probably close the captions so I don't get...
[14:18:41] I'm going to go through here and kick all these... ...and find the door... ...and the door guys... ...and the door friends...
[14:18:59] material and the finish. And then I can save that.
[14:19:16] I go back to cabinet vision and in that report, the center, I say for example is actually P.
[14:19:33] That's my number, 120769. I have to come in here, 120769. That's how the price is in here. And...
[14:19:50] you know delivery install we have a calculator you know we have we have this calculator
[14:20:08] we have to look up the distance from our shop here. We adjust the price here. Is it a remodel? Are we delivering? I mean installing and then there is another report in CalVision.
[14:20:25] I'm going to do this cab count report right here. And because I only have one race selected, I have to make sure all three are selected. It will refresh, it'll tell me there.
[14:20:42] a few cabinets, and then you know, you know, say this is 150 miles, it automatically calculates our delivery and installs.
[14:20:59] lines and side lines and what these numbers are on there. So it's a huge process. Yeah I think, well let me play this back to you to make sure that I'm
[14:21:17] You got a new job you start in cabinet vision, right? So you've designed the cabinets you place them kind of room by room Then you go up you create that report right there the you pull up the summary the report tab. Yeah
[14:21:34] You're kind of manually selecting right now which construction type and which finish level is appropriate for each room. Okay. And then like what you were saying, if the job has 20 rooms, you're making those selections 20.
[14:21:51] the more the system gets bogged down. The more the system gets bogged down, and you've gotten an increased chance of just being a human error. Well, yeah, I guess, yeah, like if one room has two different finishes, yeah.
[14:22:08] then you'd have to break. Oh, sorry, go ahead. I was gonna say, for getting the change, you wanna see some probably the biggest things. Got it, yeah, the pricing. Yeah, because you're pricing each piece separately.
[14:22:26] really you're kind of designing and pricing and cabinet vision and then you're re-entering all of that room by room data into site line right which is and I'm right like site line is wired right site line wired okay
[14:22:43] the the opportunity that the sales people are who but whatever they've already set up and then you're creating a new estimate in sightline you're picking all the relevant factors like the shift to address
[14:23:00] different options that looks like just looking at the screen. And then you're going into, was it the estimate lines and you're creating a line for each room and picking the room type meaning it? Yeah, okay.
[14:23:18] style, the construction material, all that stuff, yeah, okay. And you're doing that by looking at what can't the cabinet vision gave you. Yeah, right. So using what you call the room block, I think, as reference, yeah.
[14:23:35] Basically like you're the bridge between these two systems. The data kind of lives today in cabinet vision, but it's gotta get into sight line. And right now the only way to do it is the bridge is you. Yeah.
[14:23:52] on. Yeah, so after that, you know, it's just a matter of and I'm kind of assuming right like for complex jobs.
[14:24:09] whatever it looks like, the manual selection is just multiply. So that's that what you have to kind of keep in your head. That is that it's just a potential risk factor like for every everyone you add it just as a higher risk that maybe something isn't right.
[14:24:26] chances are that risk of getting 20 something in wrong is going to be really strong. If it doesn't take that long, maybe 5 or 10 minutes, that all those big jobs with 15 or 20 brains, I just been...
[14:24:44] I mean, I understand it, right? Yeah, like totally makes sense. How many how many of these jobs per week are you guys doing? Is it just you or is it the whole team? Oh, no, there's like five or six designers. Yep, okay.
[14:25:01] And then... I already know the one too. She just started today so she's not really doing much yet. Oh gosh, well congratulations. Thank you. Yeah, well, I mean, that's probably a good sign for the business. But, and then, okay, so when, let's say like in errors...
[14:25:18] travel.
[14:25:35] I know there are some salespeople who aren't as good at doing that. Yeah. Yeah, if a mistake makes it to the customer, like, I put in a crime from because I forgot to change my construction.
[14:25:53] going from one region to another, that can drastically affect the price. So, you know, customers looking at this, you know,
[14:26:10] right yeah okay does it ever make it to the floor or is it like by the time the customer I mean I guess it just kind of depends on the customer if they're looking at it close enough
[14:26:27] The most I do, the most my responsibility is to just make sure everything is set up correctly in CAD revision through the job and the room properties and all of this.
[14:26:45] customer goes through making changes to make sure everything gets changed because you know if you change a door or a cabinet from a regular door to a black door that puts an overwrite on the cabinet yep or on that door and so if they
[14:27:02] other door style that last door isn't going to change it's going to stay the old door style so that's part of my responsibility is to just make sure all of that gets caught and there are reports and things in captivating that we use for that.
[14:27:19] done and customers like off we hand it to drafting and they're the ones who you know finalize it get all the reports ready and everything um and create programs for the top floor and then upload it into
[14:27:36] I think I'm following that. So maybe help me map out who owns what. So the salesperson right creates the opportunity. And then you and the other designers, you guys are doing the cabinet vision design. And you're doing as well the sight line entry.
[14:27:53] And you're also kind of sounds like responsible for the initial review of the estimate before it hits the customer. That's where you were saying like that the salesman may get involved depending on who they are. Yeah, well, I mean, ideally the designer should be double checked.
[14:28:11] when you've been staring at the same job for three days, you're not gonna see everything that's wrong. And so that's why I think it's helpful when the salespeople are willing to take some time and go through it and make sure they...
[14:28:28] I mean you must have a cheat sheet just to remember what to type from one system to the other or you have an amazing memory. Yeah, my goodness, this is complicated.
[14:28:45] Yeah, okay, okay. But if you could take that out, like if there's, let's say there was like this magic bullet, right, where you could eliminate that underlying the manual work kind of going back and forth between these various
[14:29:03] from what you're saying is like if we can make it so that when you finalize and can't cabinet vision all those room details just appear in sightline automatically there's no retyping no room block that
[14:29:20] were promised that just never came to fruition.
[14:29:37] It was never really explained to me how it would be imported. Um, but yeah, that we would just be imported into the SDA that we created, and it would just, you know, magically...
[14:29:54] Macically created the estimate so you wouldn't have to fill all out and fill out the estimate like manually. Totally. Nope, makes sense. But real quick just so I make sure I have for our record, so like the designer...
[14:30:12] their work. Do you guys work as a team? Like is there a second set of eyes before the customer sees it? Or is it? Yeah okay. Yep.
[14:30:29] or if it's not like K-Designer, so salespeople have actually been doing their own designs because the client may not be able to do it all. Yeah, no, I completely see that.
[14:30:46] super interesting. I think this is something that we absolutely can explore further. There's a lot of elements of this where I think they can be automated. Again, like our point today is just really digging into each different component and figuring out what the
[14:31:03] structure is but is there maybe for you like is there any anything else maybe any other major magic wand questions like if you could fix one thing
[14:31:20] I can't think of anything else. I mean, that's the big wish is to just be able to get that data from
[14:31:38] That's super helpful. If we look at that, like how it walked me through in your in your mind how you envisioning that structure Like would you want to use I imagine that you don't want to change your work
[14:31:55] like this is muscle memory right so you've been using cabinet vision forever. I can't imagine that you have interest right now like shifting to a different tool or a different application in cabinet vision where you have to relearn everything. So walk walk me through like what your ideal vision is that
[14:32:12] you're doing that.
[14:32:30] Yeah, I'm not going to put all the stuff in for all the tabs. On the CV, it's just they feel left. Yeah, I think it's not complicated. I'm totally with you, Randy. No. Yeah. So if we already have that, it's just not a first step. That's what you're...
[14:32:47] There is a step further along when she's talking about how to practice. That's the regular step that's doing that. Okay, walking. Yeah, that's super interesting. Yeah, we'll be back. Can you walk us through what that looks like, the step that is in place?
[14:33:04] Yeah, I don't know. I just know Draft Team, when they after Genet's team and they're good and the salesperson says the customer likes the price, you know, she adds it off to the Draft Team department and they put some place like in the store.
[14:33:21] Like you ship it off, it's out at your department, and then it moves on. Yeah, they have a thing called the whistles. They put the barn app in more on to a full-on blueprint.
[14:33:39] You know, and that's all done in CME as well. Yeah, got it. Okay. That's me, Jeanette, drawing. You know, to the next level. Okay, that makes sense. So when you're designing, you know, just to dig deeper, Jeanette, when you're designing this.
[14:33:56] to change something mid-process. So maybe they're switching from like oak to maple on a bread. How much rework today does that create for you? Like do you have to redo the whole thing in? Oh no. No, okay.
[14:34:13] I think they're changing their door paint, that wood, you know, I've got all these selections. I think it's coming to the room properties right here. And let's say they're changing to a wood. So we're going to have to stay.
[14:34:30] This too, let's just profile it a little to the little bench I'll put it open and then I can roll up. So I need to include that also. And then I come in here and I change my paint grade.
[14:34:48] Black Cherry, not white, that was the wrong one. But just see how there's so many options. There's, you know, yeah. That's what we're talking about. She just has to sit and scroll through or talk to you, remember where it is.
[14:35:05] I totally see that. Let's say that was structured in a perfect way. What does perfect mean to you for a new job?
[14:35:22] these templates or Thar creating these templates. You didn't. Okay. Got it. Yeah.
[14:35:39] or like designers may create their own custom that works for them. Yeah, and I mean I can come in here and say I want a different door and I'll take the monument instead.
[14:35:56] It's... I have to change the... To that... And then I have to say there is some keyness to it even after...
[14:36:14] and change your own rule. No, it is a global issue here. There are some things where I still have to go in and change. And this is just like, you know the system, you know it's course through experience, and if you don't know...
[14:36:31] Yeah, it's honestly, it's, it's what I felt when I was going to my training designers, because that is my responsibility, is, you know, teaching them the proper way to make these changes, to save themselves time later.
[14:36:48] things that if you do them out of order, it takes you longer to make other changes later. But yeah, so I mean this is kind of tedious, but...
[14:37:05] as many as I need, time goes to stick with it. So like your new designer that just started today, the knowledge that you have, right? Like let's call it like tribal knowledge.
[14:37:23] Is that written down or how long is it going to take her or him to get up to speed to where They're they're gonna be able to cut on their own fluently move through this
[14:37:40] up probably eight years ago I think and then I updated Delsen every time we upgrade to a new cat vision version. Yeah. But yeah this is basically a step by step.
[14:37:57] a job in capital. And then you know you come all the way down here, it goes into pricing, but then for the pricing you have to jump to the why of work construction.
[14:38:14] If you don't mind in a shared drive just sharing this data for us, then we can kind of, we can synthesize it and try to figure out are there areas that we just automate. And how
[14:38:31] Like, I imagine, well I guess the, do you guys think this is a Tharp specific challenge? Because of your scale, you know, you guys are a lot bigger than a lot of cabinet makers. Or do you think this is just like the reality of doing custom...
[14:38:49] Yeah, go ahead. I think this is the reality of doing a test of cabinetry because I've actually worked for another testing cabin job that he used to have in vision. I mean, it kind of seems like cabin in vision is like the other thing, right? Like most people are using cabin in vision. Yeah. Well, actually, you know, I don't know.
[14:39:06] Interesting. Okay.
[14:39:23] after they produce. Anybody who's a distributor for a cabin manufacturer, they're just using a basic drafting software. Got it. That makes sense.
[14:39:41] How would they handle that same gap? I could probably just get the same kind of annual process. Oh, okay, so like who knows?
[14:39:58] But what we did is we would take this report, this report tab, and we would export it to Excel. And then we would have to go in and double check.
[14:40:15] I don't know if you've ever mentioned, you know, is it of the right percentage of size, or size of kids, you know, we would have to double check all that. And then, I don't know if it will be...
[14:40:32] But, yeah, it was a pretty tedious process for that too. Um... Yeah, I don't know what they're doing now, but... No, that makes sense. Um... Okay.
[14:40:50] I'm not sure if this software can set those designs and layout on cabinets. But I'm trying to put the pricing model within it as well. And you can get it, which we have a little bit too complex. So, you know, you're...
[14:41:07] versus that, instead of just sending back and really looking at it overall. Yeah, totally. You know, Randy, on that point, if we dive deeper there, like how do you see this all playing out? Like what would be your perfect setup?
[14:41:24] Um, HPP, well, my, I mean, it kind of goes back to my record. Like, I'm more of, you know, with advanced action, okay? Experience and repeat. Yeah. Where the citizens are more accustomed in, you know, that's...
[14:41:42] So it's not like every cat or cat's size, this can be 15 inches by 30 inches by 24, it's not. So and that's the world that I've come from and tried to implement all these different sizes.
[14:41:59] Europe he makes that work challenge so that's what I like to say you have to really step back slow down double check everything before you before you can you know move it
[14:42:16] they need to buy how many later feed and now we're not even talking the species of what you know no completely in the
[14:42:33] all these different variables, different finish combos, like percusp, all those different elements, those aren't going to change, which makes me wonder when you're actually, when it comes down to the pricing logic engine that you're implementing, where does that
[14:42:51] What does that live? Is it in the rate tables or in sight line and spreadsheets? Or is it the true answer there that it's kind of in your head and you know the general range just because you're expertise and then you're spinning it out?
[14:43:08] I mean, they put it in there, but... Can't control it.
[14:43:25] or something like that, company owners, what model, what's the rest of DNA, what are you looking for? Sorry, I didn't mean, like the estimate that's sent over to the customer, what is the single source of truth for that? Or what's the final?
[14:43:42] had a vision gets all the numbers is in this material manager. Okay. You know, we have all of our projects in here with their price, you know, everything is in here. And you guys are setting this or this is just like it's a flow?
[14:44:00] for it? So I believe Randy that, so we have a team of what we call systems engineers. And they're the ones who are in charge of maintaining CAP division. And I believe what happens is Randy.
[14:44:17] there are going to be price increases or price decreases. And then it goes to management and they make the decision of, if we're going to change our prices and get filled by how much.
[14:44:34] So this may flow up through Santiago or this flows up through the underlying organization. This is updated based on their analysis.
[14:44:51] You can look at those and go, you know, it's great where a person is now in that sense, you know. Okay, so maybe it went to 10 cents. You know, is it worth the change? You know, what's the bang for the buck when we kind of look through? So...
[14:45:08] that fall into like the COVID era, when everything changed, you had to cherish and reflect a lot of different stuff. You have your normal working conditions. Well, that's a great question, which is like how often are these braces getting updated?
[14:45:26] and I mean, we're supposed to update that quarterly. Yep, quarterly, okay. Well, it's basically kind of, you know, system insurance drags out, you know, and then if there's a big question from upper management saying, hey, we're seeing a lot of tariffs and pricing change.
[14:45:43] And we implement that and we decide what they're going to absorb and what they're not going to absorb. Yep. You know, it may be updates or... And I imagine there's like scenarios where there's quotes that are going out based on old pricing because the update hasn't...
[14:46:00] I can do that. So we're giving the prices to ourselves for some young, can't really adjust it. Yeah, yeah, yeah. Yeah. We gave you a price and then a while ago, six months, you know, we might have a disclaimer in there, but I think that's right.
[14:46:18] I mean how often does that happen? Is that like a major frustration point or is that more you know ad hoc?
[14:46:35] sometimes take a year to two years to build. Yeah. And they have to come to us before they even break ground to get an initial quote that they can then give to their bank to get their professional loans.
[14:46:52] I've been in the chat for a couple of years and I don't think those customers expect prices to stay the same. We do have a disclaimer on the quote that says this quote is balanced for 30 days from the date.
[14:47:09] And most people are okay with that. I think it's probably the smaller people who are coming to us just for vanity. But those kinds of goals are required. Those jobs tend to get sold within a couple of months.
[14:47:27] And there have been high sales pricers that have said, don't worry about your pricing, this is not going to change. So it's kind of up to the sales person whether or not we reprice after those 30 days.
[14:47:44] like we go at free price every job every 30 days. You know? Yeah, like also, oh, sorry, go ahead. Well, that job is, you know, it could be up every six months before the customer comes back and says, okay,
[14:48:01] to that particular price that presents Mars. Yeah, yeah, yeah. Okay, that makes sense. Okay, I think I understand the cabinet vision component. The one question on sight line,
[14:48:19] this but is there any important capability right now like a like an API or a CSV anything or they have it it just doesn't work? We just haven't implemented. You haven't implemented, okay. Yeah, we haven't.
[14:48:36] Yeah, that's what you were saying. Okay. Yeah, got it. But, you know, it's just engineering getting that set up and having the time to validate that and make sure you know what I mean.
[14:48:54] answered this but I'm scribbling down notes I just want to make sure I've got all this all this accurate it how many jobs like how many total jobs per week across all all six of the designers would you say just generally how many you're doing
[14:49:11] Um, I would say probably anywhere from 20 to 30 in a week. Got it, okay. Um, okay, this is super interesting.
[14:49:29] a very complex, incredible organization and operation over there. Well, I guess broadly, when you look at your current work that you're doing, what do you have the most?
[14:49:46] matter like this this is where we just excel or this is where I individually and just really good
[14:50:03] The jobs that I have worked on very rarely will have design skills related to your orders or non-follables that have to process after you stop. Yup, so accuracy, quality, that makes total sense.
[14:50:20] the fact that I've been using cannabis for almost 20 years, is that I know how to make things look really good. And I know what the limits are of what we can do.
[14:50:38] You know, I'm not going out there to be honest. You can't imagine you can make something look right. Yeah, but can you actually make it into the real world? But we can't actually make it that way. So, you know, it's all the way.
[14:50:55] very good understanding of what we can and cannot build and how we can build things. So I'm very good at making sure that what I'm presenting to the customer in 12.
[14:51:13] about that. And it probably goes back, like, does that kind of go back to the documentation that you created of, like, these are the major, okay. Yeah, like, that would be super helpful for us to see as well as, is it possible for us to get sample
[14:51:30] maybe like a sample of what sight line expects on the import side. So we can have our engineers kind of play around with it. Yeah. The only way I was looking at this job, I just created the old.
[14:51:47] export type I had is what you call CSV/Excel and I have no idea what that is to me. Oh that's fine just send us that we'll figure it out. Okay.
[14:52:04] else on your end? What I'm kind of hearing from you, Randy, is you come from a broader background in terms of just broad manufacturing. It'd be awesome if you could have structure and car rails, right? And if you can just leave
[14:52:21] having to put in that's high stress mentally. So you can focus on like the higher margin or the higher utility without just going back and forth and trying to check a bunch of boxes.
[14:52:39] And I was trying to tell people what's the second word in our name. Yep. You know, and really from that year on, we're not doing this repeat. Yeah. And you know, somebody comes in with a crazy idea. We should be able to...
[14:52:56] No game on I love that I love that idea Yep
[14:53:13] That's the timing that I'll call it. The in-between the home depot with lows and fully-dusted scapula. Yep, yep. So, you know, and then price it more, but to have that for instance, repeat a limited option.
[14:53:32] the cost model is in line. But it's hard for them to manufacture it with the production line. Production doesn't know the difference. You know, they see a job and they stand and stand it.
[14:53:49] We don't know this job, this Virginette's million dollar customer up in the mail. That's going to be super picky. Yeah, that's an interesting point. It'd be nice if you were able to kind of parse those things out. Yeah. Yep. And we try to pry it with the usual...
[14:54:06] build cycle or certain level build cycles. And that's kind of the thing. I look at product manufacturing versus full-growth manufacturing. Yeah. You know, that's the team.
[14:54:24] on the floor, that way you're trying to run everything the same. And that veil comes from one floor, tension is detailed. And, you know, the track homes, you know, might not.
[14:54:41] No, 100%. 100%. Yeah.
[14:54:58] For me, it's trying to simplify the options as much as possible. And that's what Jannette's going to be back to say. She knows what the customer car wants, she knows what we can and cannot do, and sells great.
[14:55:16] total sense. In Genet, like when you're building up that shared folder, anything, don't worry about customizing anything, like just give us a data dump of anything that you think could be relevant and then what are the things that you think could be relevant.
[14:55:33] it. It is Santiago probably told you guys we've got a number of different sessions lined up this week so like tomorrow we're meeting with Sam on production scheduling and then Wednesday we're going to talk to Chris on project management.
[14:55:50] further dedicated sessions post that. Right now all it is is like this initial data dump, understand what it is that you guys are experiencing so we can try to figure out a manner to synthesize it all together in something.
[14:56:08] help people and not create more confusion and more complexity. So we're in the early, early stages of that, but everything that what you've shared is super, super helpful for us to begin that puzzle.
[14:56:25] Well, that's your choice. If it is easier on your end, we can send you out. Do you guys use the Google Suite by Google Drive? I do. You do?
[14:56:42] share, drive on RN and then we'll shoot you over an invite, Jeanette, but you can just upload things, maybe very straightforward. Okay. Perfect.
[14:56:59] Anything else top of mind anything else that this conversation is scratched for you?
[14:57:17] I know very little about character vision, so I'm on the opposite side of the spectrum and I don't have to deal with it with that respect. You know, I basically give my data from what information we have inside mine, or you know, full packages.
[14:57:34] just the mention here and then they put it in, you know, they put in the dollar, they put in the percent markup, they put in the yield factor, you know, all the things that are going to be, you know, kind of external to myself and...
[14:57:51] different you know a bit total sense yeah yeah okay this is again super helpful guys kind of a
[14:58:08] So, our next steps, we'll send you over that invite, Jeanette. Again, just drop in whatever you're comfortable sharing and then we'll map this against the other sessions this week. And then we'll report back to you what we find.
[14:58:25] Alright cheers, thanks again for taking the time. Yeah, thank you. Alright, see you guys. Bye. I mean, maybe? No idea.
[14:58:43] నినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినిన
[14:59:01] I mean maybe it's so hard to tell Well, okay, you know what I think is gonna be the huge advantage of this is that we're gonna have this like
[14:59:18] a whole bunch of random shit. Like I had no idea what we just talked about in the other ones bro. But I feel like we can map it. Like I feel like I was like understanding what she was doing and kind of mapping the different pieces.
[14:59:35] But I feel like I feel like then maybe you connect ten of these and then all of a sudden you're gonna see something But then maybe we identify as but no way you're doing that. It's like the sheer information shares
[14:59:52] and then you can actually really build that and exponentially help out what Janeth is doing. Well, and then, yeah, to your question though, is that does it just help Janeth, or does it help, like, others in this industry? Right, this was the exact...
[15:00:10] I mean, that piece is confusing to me.
[15:00:27] Capitum vision and the swipes or whatever. Right? It also doesn't seem, it kind of sounds like they just haven't built that internally. Yeah, period.
[15:00:44] Yeah, well that was like, alright, we can't re-tilt this if it's a real thing. But, I mean, it fits so well, Tyler, with the broader thing we're hitting with all these companies, which basically the thesis is very simple, which is...
[15:01:02] systems that not have them talk to each other. Like that's it. So if you want to be like 95% of this conversation through a lot of discussion. Yeah, so it's figuring out. Yeah, like it's figuring out, which is what HubSpot figured out 20 years ago. And they're like, okay, we're gonna build.
[15:01:19] them but then you have to have people that are building them and then nobody's actually using it. It's like can we figure out that middle layer that right like that you're able to actually make those connections and you're not a consulting company going in.
[15:01:36] and then move on. I don't know. That's what we still know. We have no idea. No idea. Well, I mean, I also kind of think like hearing this, right? So this is just the general systems integration play. It's not replacing...
[15:01:54] and hell those six designers are gonna look she knows cabinet vision you're not changing that you're not building that you're not like we're where Marshall works out right like what Marshall what we are doing is replacing an actual human
[15:02:11] about these common integrations. Maybe there's an easy way in which you can somehow alleviate that, but it's for the next five of these that we have, like trying to find the engraving workflow.
[15:02:28] Yeah, yeah, I actually had spinny try thinking about that. I'm really curious to see her, um, that step-by-step long document she has. Oh, you know exactly what that is.
[15:02:46] Like that thing, 10 years ago, this to me sounds like when we first were like building out the research investment process for Cook Street and we found investment managers, but I think this is going to be
[15:03:03] And there's no chance that she never actually pulls that stuff out. I mean, but it also is interesting, right? Like, part of these conversations are, it's navigating, it's navigating the element where she's like trying to justify like...
[15:03:20] Right? Like it's like not upsetting the Alpacard as she did with what she actually is doing to the part where she shuts down and being like, hey, we're supporting that. But there's no way. That's quite well. Yeah, there's no fucking way that whatever's on that word doc is...
[15:03:37] was eight years ago. Yeah, I agree. Um, okay, so... I mean, I don't think there's any... Wait, go ahead.
[15:03:55] I think it 100% could be, but I don't think we do anything until we have all five conversations. Like I think all we're doing right now is this initial data poll.
[15:04:12] from this call is send her out to share Google Drive and then we should structure it such that like theoretically we're going to get this from each of the five different departments for the four different departments. So we can build out like that.
[15:04:29] But I guess it'd be like customer materials. Yeah, exactly. I need to build that right now. Yeah. And then I think, honestly, like I think it's after we have five of these, which are gonna sound very similar, but.
[15:04:47] thread that we can find. And the thread that would be surprised if the thread for like there's no, my girl and I were talking about we then send it off to so leave the designers and it goes to what's the next department? It was the...
[15:05:04] And Randy was talking about like they have this all figured out. There's no fucking chance they have that figure out. Like it just means they're desk. Yeah. So I think what, like my gut is we're gonna keep encountering the exact same thing.
[15:05:21] department that you talk to the the unique thing is that their systems don't talk to each other in which case you have to figure that out or or there's still this like there's this do you find
[15:05:38] flow that's taking somebody 20 hours that's high value that just you you fix that it's not so inter system dependent I don't know the answer to that well I know the former is gonna happen over
[15:05:56] I don't see any quote with that. Everyone's trying to figure it out. Yeah, you know, everyone's trying to figure it out or it's already been figured out and it's just not working properly like it is with this situation, right? Well, you could be.
[15:06:13] right? No and there they're without question is like an absolute an absolute moat in business surrounding if you went into Tharp and like there's no question that James could program that. Right like make that like
[15:06:30] If you're kind of building this out, then you charge them $50,000 for it. You could see doing that a thousand times. But the question is, does that... it doesn't align with our original...
[15:06:48] out. I don't know. I don't know. I know exactly what you're saying. Right? Like you like what's the actual vote? Like every every two weeks it seems there's there's something new that is just obliterated.
[15:07:05] be obliterated. This will never be obliterated because no matter what comes out technology-wise, far so good a degree in the way they are, it comes in and fixes it. Million percent. I think I've worded that before.
[15:07:22] not going to be obliterated for a long time is the actual company by company integration. Yeah. Which is what we're trying to avoid, but is that a full circle? I don't know. Look, there's... I just... Do you feel any reason...
[15:07:40] Right, but yes, but also like do I think that we could get a contract with Thart for 50 or 100,000? Yes Yeah, you're like great news. What are you paying?
[15:07:57] You don't need that. Yeah, I mean, okay Like that feels real right So maybe that's the right okay, so that that gives us good context for conversations tomorrow
[15:08:14] just keep finding that there's like there's ten of these different scenarios where there's probably an extra guy or multiple that you don't need and you you come in and you maybe maybe
[15:08:31] I mean, there's no question there's a market for it. And there's no question that you can...
[15:08:49] Randy, when you see the interaction between those two systems, there's nothing complicated about that. No, no shot. No, that is pure inputs on one side, exports on the other, you have a script.
[15:09:06] Today, they must have some sort of IT department that can't be this big, right? I think they do, but I just don't think people are... I don't... But people are... Is that... are we talking something like that? Um... Here, I'm gonna put the papers.
[15:09:23] I mean, I don't... It looks like no. We're talking...
[15:09:40] Ah, shit. Let's go to the police. Okay. Okay, sweet. Uh... Just like, you ain't from there.
[15:09:57] It could be so obvious to me to bring in an IT guy and they may seem like us but just have someone in total that's just kind of tickering with all this stuff. Like build these things out, yeah. And yet somehow...
[15:10:15] I know people like you couldn't say the same thing about our old company, right? You have like 35 people. We have all these disparate systems. It just never is like an actual priority for people. Yeah, but I don't have a good answer for why.
[15:10:32] to find something that goes female. Okay, so we're talking to... We have project management, which is Chris over... Productions.
[15:10:49] Yeah, okay. But that, like that element, maybe Tyler it's like, let's just have all these calls. You definitely can see this scenario where...
[15:11:06] you hire James, right, to just come in and like just build that system, okay, you're gonna have to pay him $100,000 a year or like probably more than that. You know, on that front, do you start building out?
[15:11:24] consulting arrangements where we've got five of them, ten of them, and you're making 500 grand. This is just so terrible.
[15:11:41] that uses some new challenges, but that would be the path that leads to resistance. Well, I also don't know if it actually takes, like, okay, think of our actual problems that we have today, which is that Aaron and James are right now in the play in absolute...
[15:11:58] the parent with weight. Right, so he's completely distracted. And then James all of a sudden is like, if we just have this, like if we prioritize trying to get...
[15:12:15] such that you take on those mandates and we still are pursuing these other components. Maybe not. Isn't that the worst idea? That's right. The actual pain point that we're experiencing right now is not that Jane...
[15:12:33] doing this but you got kicked out of his office. That really, really sucks. No, and like I can't tell you how big the impact is on Aaron. Like Aaron just goes dark because he's just like it's too.
[15:12:50] Which then all that does is set back all these other elements. So if you if you had this runway Let's think about that I mean, I'm not I'm not totally it could be after going through this entire thing. We come back to guard
[15:13:07] this like we don't see being able to sell this to other cabinet makers but we do see a shit ton of problems in your existing structure and we've got like we're identifying like two million dollars of potential efficiencies
[15:13:25] a 25 grand a month retainer. It just start working through them. Like, I'm feeling like I'm gonna be successful. Yeah, no, I thought I saw a lot of negative goods at the end. Yeah.
[15:13:42] they're still gonna be something there. I don't know, 100% they're gonna be there. It could be that literally every single, I mean, these guys have no idea. Like, some of the questions I asked them aren't even fair where I'm asking them like specifics about what they're competitive. They don't do that.
[15:13:59] That's not their role. I have a feeling if you actually are able to synthesize all these different conversations, we can figure out pretty quickly, is there, like, is the cabinet vision being or what
[15:14:17] But like is that just an endemic problem where you build a wrapper? But all we're actually doing if you think about it with Marshall is we're building a wrapper on top of the epilogue prayer. Right? Yeah. Like we're not like all of the... Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah.
[15:14:34] Like they heard that idea and now they're gonna go try and build it on their own. And we'll see if they're successful or not. The problem with the actually express model is like they now know we're doing that and they're trying to replicate it. And if they're gonna be successful or not, we don't know.
[15:14:51] You know, like it's probably not that far away to where they are to either build like this this supermode But if if they're coming out and they're just everything that we produce right they just they're They're just like sweet
[15:15:09] I mean, I could see an issue with that. Yeah, you can run into that. It varies with the time. Yeah, totally. I mean, now that I even articulated that, I feel like that entire epilogue call was just for pucking.
[15:15:26] And they're like, we're gonna go try to do this internally for the next few months and then we can't. We'll call you back. Yeah. I mean, you could see, can't have this, you could see anything. Anything, you look at every business right there, and you look at sales, like.
[15:15:43] Every single one of them, their stock is down 50 to 70% because that means it's assuming people are going to do it on their own. Like that's the biggest component we're going to keep running in circles on with what we're trying to do. But the one. Yeah, the one.
[15:16:00] But, but like the reverse of that is well if you're gonna have like 98% of companies need to implement this and if you could become a just extraordinary implementer, you know, I mean that doesn't...
[15:16:18] Yeah, you just have to go a slightly different business. I mean, I don't even know, like how different, well it's different, but how different is it than what we're trying to do right now?
[15:16:35] You only have what walked me through that. Well, the difference is if this is the goal number one is to just be like an awesome, like a class while we're in...
[15:16:52] All you need is just a bunch of people who could go in there and solve problems. There's no kind of strategy around bringing a product out to the market and selling it. Well, except for the fact that the product that you're bringing is that you...
[15:17:09] So in our way, like all these different components where we hear the problem we just heard and it takes it takes a week to deploy it effectively or less. Like on your actual system. Right.
[15:17:27] building that out. There's no question. Well, I don't think there's a question. It'd be interesting to pose this to James and Aaron. Like just hearing that one specific pain point, but how long would it take to make an integrator reach?
[15:17:44] I've been feeling like I can probably do that tonight. Yeah, and then you go to the landscape versus the exact same problem. Yeah, just two different pieces of software. But like the infrastructure of it is not that different. I mean, time is not. Like that's.
[15:18:01] for them super quickly you don't share it no the other big guys are gonna roughly or two again three months you just deploy it $8 yeah yeah and you look at it and like after today okay so we don't know how much they pay desires but we can find that
[15:18:19] like pretty quickly that she's spending four or five hours doing these, I forget how many jobs she said a week, right? Whatever it is, if you come in and they're like, okay, well, you're going to spend, you're going to spend 500,000 on humor.
[15:18:36] It seems like 20% of that is trying to manage the complexity of this system. So we'll build you something in the next two weeks that fixes it and we'll do it for 50 grand.
[15:18:53] six months when every single company on the plant starts firing people and starts doing this instead. Yes, and then it's like a 30-hour- Yeah, like it's 100% happening. And if we just, if we focus at our energy on being like-
[15:19:11] that solution. And you also then are beholden to, okay, we put all our work in this one workflow and then the laser manufacturer is offering it for free. Which for sure is happening.
[15:19:28] The moat we're trying to build over there is that it's like the trust network with Marshall where you guys are in together and it's like this. And I think that's real. I think that definitely is real to a degree, but is it easier just to be...
[15:19:45] I call it execution on a specific problem and then move on to the next one. Yeah and it would be nice to have like maybe something in the middle where you're not just doing a pure consulting plan or you come up with something like...
[15:20:02] And that was like 75% done. And then customized for last 25% for whoever added fee if they don't want to take it out of the box. So you're not kind of- Oh, that's interesting. Yeah, so it can be like, or you can either do this for a thousand bucks.
[15:20:20] like upfront fee if you'd rather do that. Is that where you think it's going to be? Yeah, because then you're bringing kind of like a half big product to someone. Yeah, and so that may have a longer... This is a little like a paltier, right?
[15:20:37] about okay so I kind of love this idea if you think about what it is all the things that we're building internally if let's just say candlefish is the product right and everything we do all we're trying to do is get better at
[15:20:54] is deploying something that works quickly and effectively. And then you're like, there's the multitude of different ways you can pay for that. You can do a subscription model. You can do, but it's almost changing from, I mean, that kind of isn't.
[15:21:12] There's more in bridging than has some cabinet industry rates. You still spend, right? I don't know, maybe a couple weeks a month of R&D on your first project, call the spy or whatever. Yeah. And then you have that. And then you could quickly go.
[15:21:29] at the time from the start from the basic kind of foundation that we realized is where people are having most their issues. It's never going to be the same every time which we're coming to learn. No, no. And also like you don't know it's
[15:21:46] no longer his ballot. But and then doesn't it also open up so like think of Sonamow like instead of like he's all in on that concept and he's worried about what this means for the long-term future of his daughter. If instead the model is like no you bring us
[15:22:04] we're going to analyze your systems, we're going to identify something, we're going to say we're going to be able to like collectively look at it and say this right now you're spending 100,000 or a million dollars on this, we're willing to tackle that problem in a bit.
[15:22:21] a 20% one-time fee or you can sign up for a two-year guaranteed subscription model or a fraction of that or whatever it is. Yeah. Like just even turn to a conversion as opposed to
[15:22:38] Yeah, and you never just do like a one-off, right? No. You still use that and keep working on similar issues with similar businesses. A million percent, yeah. And you try to promise the same. Yeah, I think we have to be careful.
[15:22:55] for, I mean, the Unburdened guys are gonna be kind of interesting to talk to about this component because I don't, I think you're, I think you're a little bit screwed if, let's say you sign people up on a retainer for 10 grand a month.
[15:23:13] But if it's like all of a sudden, then they're expecting, they're like, no, we're paying you. You guys work full time for us. It's like this third week, right? They're like, yeah, six hours of meetings. Right? So it's like, almost you want.
[15:23:30] intake and then identify things that you want to work on and then build out an olacart pricing model, right? Which is... and I think you could do it still. I think the development can happen quick enough if we find things that...
[15:23:48] know we just have better app right now. It's information and asymmetry. And if you're like, okay, we'll do this for 20 grand, it'll be delivered to you within three or four weeks, or you can do a subscription model. And then we can also like start there and then you pick that one off.
[15:24:05] You get a discount. Yeah, I like that. It seems like this might naturally grab a take towards that. It almost seems like it could be the only model right now that's in between.
[15:24:22] companies right now that are trying to build software that then... I guess the way I'm thinking about this is so we're trying to build a SaaS company right in the midst of an environment where all the all the real SaaS companies are
[15:24:39] For one reason, because they think that every company is just going to build their own tools. Which is exactly what we're doing. Yeah, exactly. So is the real utility the real value? I mean...
[15:24:57] And it solves some of our other actual shorter term pains. Right, like I said, like clipping off like 200 bucks, 400 bucks a month, you start clipping off 10 to 20 grand or whatever. I mean, I think we price this in so many different ways.
[15:25:14] quickly get to the point where you're like okay we got 300 grand in non-recurring software revenue but what that does is then we have you we have James we have Max right? Yeah, this is quite a tight time.
[15:25:31] action. Yeah, well, maybe that's the way to phrase this is this is the fastest way. Yeah, like what are we actually trying to solve for to get us whole? Yeah, so we can actually figure out what we're gonna do the next few years. Yeah, yeah, because what
[15:25:49] is we have these ideas and then reality hits where I've got a week like this week, right? Where I just did four fucking meetings and then you're jumping on this, right? Like that, that just is the practicality for a period of time. But then you've got.
[15:26:06] like he's trying to, like he's disconnected. But James, you know, he's trying to keep his current job not outside the Alpacard. They're like too many distractions. Yeah, and everyone's just waiting for a product to go market and see.
[15:26:23] to ramp up. Yeah. And then you're like, okay, but also you're like breaking even. Like, what? I mean, for, for, I don't care about the money part, the more is like, how do you keep it sustainable? Right? For like, keep it sustainable for like
[15:26:40] you pivot, right? But people's attention and focus. Okay, I, I kind of really like this idea. Yeah, I did too. One thing we know we have is we should finish these next four calls. No, yeah.
[15:26:58] I think it helps framing because I don't I don't I don't even know I mean we could pivot this issue After all these calls we could come back to third First of all we may find something that we're just like it's the takeoff estimator Or if you figure that out like it's worth trying
[15:27:15] But if it's not, then we come back to him and we say, look, at the end of the day, we can't find one thing, but you guys are right when you first analyze this. This is probably why his body went into an enterprise. Which is like, if they...
[15:27:32] I was really excited about it. But you look at this and you're like, okay, yes, you guys have a ton of problems in the hallway, but it seems a little bit complicated to transform 8,000 others that are gonna all be unique in the room, right? But we also kind of...
[15:27:50] that we think you guys don't need to pay anymore. Yeah, well, if you're gonna do this, it does make sense to your enterprise, right? You find the money that's gonna pay a billion dollars, not 10,000 dollars. 200%. For the same amount of work. You know, related to this,
[15:28:07] I'm looking forward to this. Marco Sents, he's so funny by the way. But I had a conversation with him yesterday where he, like he's just like in a very friendly way. He's like, can I get involved in every dark?
[15:28:24] with him and came up with he's got a buddy back he stopped it I just ordered to you I looked at it but that could be that could be this discussion point we have others like we have others where if you shift the initial discussion
[15:28:42] anything up front but the intention is we're going to identify that you're... instead we're gonna identify something that we then can go co-revenue sponsor. Maybe we try a few of these. We're gonna identify something where you...
[15:28:59] historically that we think we can solve and we'll solve it for 20% of what you're paying in a normal year. Yeah and if there's ten things maybe the first one for free.
[15:29:16] Yeah, do the first one prove that works. We're gonna find 50 things and we'll say, alright, we're just showing you in two weeks and then if you like it, we'll do the next toy for you. Yeah, it'll take us something. I don't even mind...
[15:29:34] Like it's but I also think it's sort of like if if we build this and you use it then you're paying us for the first one Or maybe you do what you're saying, which is like the first one's free so we identified this and then they're just like that is fucking
[15:29:51] That just depends on you know, yeah circumstances Yeah, until until we yeah until we have a reputation where people are
[15:30:09] forms by systems just fucking pay them whatever you you do that right you build the reputation of it but like our our capital right now is that we've got a lot of people they're like super interested and I've no idea
[15:30:26] Let's build a product for the future that nobody knows what the future is. Maybe just solve the problem now
[15:30:43] Okay, keep going. Okay, bye.
[15:31:00] Thank you. Thank you.
[15:31:20] ლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლ�
[15:31:55] ლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლლ�
[15:32:13] నినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినినిన
[15:32:31] I'm not sure what to do with this. I'm just going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office. I'm going to go back to the office.
[15:33:06] සිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිසිස
[15:33:24] වවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවව
[15:33:42] සිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිවිව
[15:34:00] වවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවවව
wake/logs/mcp-stack./INTERNAL_TEAM_HANDOFF_2026-03-04.md ./INVENTORY.txt ./core/THARP_72H_MACSTUDIO_GATEWAY_INTEGRATION_2026-03-03.md ./core/THARP_CHRIS_PM_LIVE_CALL_SHEET_2026-03-04.md ./core/THARP_CV_WIAT_BOARD_BRIEF_2026-03-02.md ./core/THARP_CV_WIAT_EXEC_SLIDES_2026-03-02.md ./core/THARP_DEEP_DIVE_AND_TOMORROW_PLAN_2026-03-03.md ./core/THARP_EMAIL_BODY_BEAUTIFUL_2026-03-04.txt ./core/THARP_EXEC_DECISION_MEMO_FOR_SANTIAGO_GARTH_2026-03-04.md ./core/THARP_TOMORROW_DEEP_DISCOVERY_PROMPT_2026-03-04.md ./meeting-intel/README.md ./meeting-intel/calls/2026-03-02-ms-imr-bn__summary.md ./meeting-intel/calls/2026-03-02-ms-imr-bn__transcript.md ./meeting-intel/calls/2026-03-02-tharp-cv-pricing__coaching.md ./meeting-intel/calls/2026-03-02-tharp-cv-pricing__summary.md ./meeting-intel/calls/2026-03-02-tharp-cv-pricing__transcript.md ./meeting-intel/calls/2026-03-02-tharp-debrief-tyler__summary.md ./meeting-intel/calls/2026-03-02-tharp-debrief-tyler__transcript.md ./meeting-intel/calls/2026-03-03-tharp-debrief-internal__summary.md ./meeting-intel/calls/2026-03-03-tharp-debrief-internal__transcript.md ./meeting-intel/calls/2026-03-03-tharp-production__coaching.md ./meeting-intel/calls/2026-03-03-tharp-production__one-pager.md ./meeting-intel/calls/2026-03-03-tharp-production__prep.md ./meeting-intel/calls/2026-03-03-tharp-production__summary.md ./meeting-intel/calls/2026-03-03-tharp-production__transcript.md ./meeting-intel/calls/2026-03-04-tharp-cv-wyatt__prep.md ./meeting-intel/calls/2026-03-04-tharp-pm-workflow__one-pager.md ./meeting-intel/calls/2026-03-04-tharp-pm-workflow__prep.md ./meeting-intel/coaching-protocol.md ./meeting-intel/followup_Candlefish Internal Weekly Call_20260303_1802.md ./meeting-intel/tharp-coaching-full.md ./meeting-intel/tharp-consolidated-internal.md ./meeting-intel/tharp-cross-synthesis.md ./meeting-intel/tharp-raw-full.md ./ops/OPS_STATUS_SUMMARY_2026-03-04.md ./ops/apple-notes-mcp.events.redacted.log ./ops/gws-mcp.events.redacted.log ./ops/readai-mcp.events.redacted.log ./ops/wake-mcp.events.redacted.log
./INTERNAL_TEAM_HANDOFF_2026-03-04.md ./INVENTORY.txt ./THARP_COMPILED_MASTER_BRIEF_2026-03-04.md ./core/THARP_72H_MACSTUDIO_GATEWAY_INTEGRATION_2026-03-03.md ./core/THARP_CHRIS_PM_LIVE_CALL_SHEET_2026-03-04.md ./core/THARP_CV_WIAT_BOARD_BRIEF_2026-03-02.md ./core/THARP_CV_WIAT_EXEC_SLIDES_2026-03-02.md ./core/THARP_DEEP_DIVE_AND_TOMORROW_PLAN_2026-03-03.md ./core/THARP_EMAIL_BODY_BEAUTIFUL_2026-03-04.txt ./core/THARP_EXEC_DECISION_MEMO_FOR_SANTIAGO_GARTH_2026-03-04.md ./core/THARP_TOMORROW_DEEP_DISCOVERY_PROMPT_2026-03-04.md ./meeting-intel/README.md ./meeting-intel/calls/2026-03-02-ms-imr-bn__summary.md ./meeting-intel/calls/2026-03-02-ms-imr-bn__transcript.md ./meeting-intel/calls/2026-03-02-tharp-cv-pricing__coaching.md ./meeting-intel/calls/2026-03-02-tharp-cv-pricing__summary.md ./meeting-intel/calls/2026-03-02-tharp-cv-pricing__transcript.md ./meeting-intel/calls/2026-03-02-tharp-debrief-tyler__summary.md ./meeting-intel/calls/2026-03-02-tharp-debrief-tyler__transcript.md ./meeting-intel/calls/2026-03-03-tharp-debrief-internal__summary.md ./meeting-intel/calls/2026-03-03-tharp-debrief-internal__transcript.md ./meeting-intel/calls/2026-03-03-tharp-production__coaching.md ./meeting-intel/calls/2026-03-03-tharp-production__one-pager.md ./meeting-intel/calls/2026-03-03-tharp-production__prep.md ./meeting-intel/calls/2026-03-03-tharp-production__summary.md ./meeting-intel/calls/2026-03-03-tharp-production__transcript.md ./meeting-intel/calls/2026-03-04-tharp-cv-wyatt__prep.md ./meeting-intel/calls/2026-03-04-tharp-pm-workflow__one-pager.md ./meeting-intel/calls/2026-03-04-tharp-pm-workflow__prep.md ./meeting-intel/coaching-protocol.md ./meeting-intel/followup_Candlefish Internal Weekly Call_20260303_1802.md ./meeting-intel/tharp-coaching-full.md ./meeting-intel/tharp-consolidated-internal.md ./meeting-intel/tharp-cross-synthesis.md ./meeting-intel/tharp-raw-full.md ./ops/OPS_STATUS_SUMMARY_2026-03-04.md ./ops/apple-notes-mcp.events.redacted.log ./ops/gws-mcp.events.redacted.log ./ops/readai-mcp.events.redacted.log ./ops/wake-mcp.events.redacted.log