Mario Software (2D/3D Modelling Tools) concept partnership
Help Mario build AI-assisted 3D modelling from 2D drawings + reporting system. Blocked on inputs: his two framework documents (private Google Docs) and an agreed deal structure. Time-boxed: docs + structure by early October or park.
Why we hold it: Leverage your AI/automation build capability onto Mario's process-mapping methodology. The technical core (2D→3D) is genuinely hard, so the value is in choosing a narrow MVP slice and structuring the deal (paid spike first) rather than in building immediately.
Docs read 2026-09-15: ambition is full-auto 2D→3D, zero demand evidence — scope conversation + written deal by 2026-10-03, or park.
✅ Proceed when
Docs + structure + funded spike agreed — then treat as active build bet
⏸ Park when
No docs/structure agreement by 2026-10-03 — park without guilt; the idea keeps
⛔ Kill when
Docs reveal research-grade ambition with no funding, or IP terms unacceptable
💡 Advisor: 2026-09-12: Fastest unblocking sequence in the whole portfolio: one export session + one conversation. Until both happen, spend zero build hours. Authoritative registers: projects/Mario Software/03_Registers/.
🗺️ Engagement Flow — Stages, Gate & Exits
⚠️ What could sink it — top risks
RSK-01Technical feasibility overpromised — full automatic 2D→3D reconstruction is research-grade.high 16RSK-02Deal structure left undefined while work begins — unpriced hours accumulate silently.high 12RSK-03IP ownership ambiguity over the code Zaid contributes.high 12✅ What to do next
ACT-00260-minute scope conversation with Mario (user, drawing types, funding, expectations)in 3d (2026-10-03)ACT-003Decide + write down deal structure: paid vs partnershipin 3d (2026-10-03)ACT-006Define the MVP slice with Mario — candidate from his own roadmap: single-storey floor plan → basic 3D → BoQ extract (ACT-004 in project register)2026-10-10Gate criteria — Concept → Validation
- Both documents read and re-assessment updated
- Scope conversation held
- Deal structure agreed in writing
- MVP slice defined (one drawing family → one report)
Unsigned deal + research-grade technical ambition against a buildable PoC slice: build nothing until structure and one narrow slice are pinned. · Decide with evidence whether Mario's methodology and Zaid's build capability form a fundable collaboration, and on what written terms — converting the engagement into structured paid work (feasibility spike first) or parking it cleanly by 2026-10-03.
🗣 Problem Statement
Mario's vision documents (read in full 2026-09-15) specify a fully automatic 2D→3D BIM pipeline — research-grade ambition with zero demand evidence — and a 5–6 person RACI for a team that doesn't exist. The buildable slice is Mario's own PoC: single-storey floor plan → basic 3D. What blocks progress is no longer evidence but the deal: no scope conversation held, no structure written, no slice agreed.
Pain points (severity)
Evidence & sources
- Two framework documents (~11–12 KB each, last edited 2026-09-11) exported to 04_Inputs and read in full 2026-09-15; re-assessment §8 appended
- Deal structure undecided — DEC-001 parked with weights fixed (cash-flow 0.30, upside 0.25, risk 0.30, relationship 0.15); advisor recommendation is a paid feasibility spike (5–20 h) under any structure
- DEC-002 provisional scoring: parametric core + AI-assist 4.3 vs pure-ML 2.4 — full automatic 2D→3D is research-grade (RSK-01)
- Portfolio strategy §2: Mario unblocks with one export + one conversation; advisor line is 'paid feasibility spike or nothing'
🎯 Purpose & Objectives
Decide with evidence whether Mario's methodology and Zaid's build capability form a fundable collaboration, and on what written terms — converting the engagement into structured paid work (feasibility spike first) or parking it cleanly by 2026-10-03.
Objectives
📐 Scope
This stage covers evidence-gathering and deal-shaping only: export and read the documents, one scope conversation, a written structure, an MVP-slice definition and a paid spike proposal. No tooling is built, no IP-bearing code written, no hosting committed before the structure is signed.
✅ In scope
- Export + ingestion of the two framework documents (ACT-001)
- Full advisor re-assessment against the actual documents
- 60-minute scope conversation: user, drawing types, funding, expectations (ACT-002)
- Deal-structure decision, in writing (DEC-001 / ACT-003)
- MVP slice definition: one drawing family → parametric/AI-assisted 3D → one automated report (ACT-004)
- Paid feasibility-spike proposal and sale, 5–20 h (ACT-005)
⛔ Out of scope
- Building any tooling before the structure is agreed in writing
- Partnership or legal formation before written IP and exit terms
- Full automatic 2D→3D reconstruction (research-grade — lowest-scored option in DEC-002)
- Hosting, model-API or infrastructure spend
- Market and go-to-market work (no buyer is named yet)
Constraints
Pair Zaid's AI/automation build capability with Mario's process-mapping methodology. Because the technical core (2D→3D) is genuinely hard, the value is captured by choosing a narrow MVP slice and structuring the deal — paid feasibility spike first — rather than by building immediately.
💎 Value Proposition
Pair Zaid's AI/automation build capability with Mario's process-mapping methodology. Because the technical core (2D→3D) is genuinely hard, the value is captured by choosing a narrow MVP slice and structuring the deal — paid feasibility spike first — rather than by building immediately.
Value drivers
Feasibility spike — first priced revenue
Paid development track (Option A)
Partnership upside (Option B)
Reusable pipeline components
Revenue model: Undecided (DEC-001): (a) paid development per milestone, (b) partnership equity/revenue share, or (c) hybrid paid build + small equity kicker. Scoring deferred until documents and conversation land; the spike is paid under every option.
⚖️ Cost–Benefit Analysis
Costs to date: 1 founder hour, $0 cash; maximum pre-structure exposure is the 20-hour cap. Benefits are placeholders until the deal structure is agreed — the paid spike (5–20 h) is designed to be the first priced revenue.
Costs vs benefits (scaled)
Method & assumptions
All benefit amounts are placeholders (0): no revenue figure exists in any source and none is invented here. If the spike needs hosting or model APIs, costs get quantified at spike scoping.
Evidence-first, spike-first: read the documents, hold one conversation, write the structure, choose one narrow slice — then sell a paid feasibility spike. No build hours before all four. If the inputs haven't landed by 2026-10-03, park without guilt: the idea keeps, the hours don't. Under any structure the spike is paid work, so the first hour after the gate is a priced hour.
🧭 High-Level Strategy
Evidence-first, spike-first: read the documents, hold one conversation, write the structure, choose one narrow slice — then sell a paid feasibility spike. No build hours before all four. If the inputs haven't landed by 2026-10-03, park without guilt: the idea keeps, the hours don't. Under any structure the spike is paid work, so the first hour after the gate is a priced hour.
📍 Positioning
Undetermined until the documents identify the target user, domain and drawing types. Interim posture: senior AI/automation capability applied to a partner's proven methodology, sold first as a priced 5–20 h feasibility spike.
🏰 Moat / defensibility
None today. Candidate moat if it proceeds: Mario's process-mapping methodology codified into the product — currently the single point of failure (RSK-05), not yet an asset.
Strategic options
Paid feasibility spike first (5–20 h) before any structure commitment adopted
Priced evidence under any structure; costs one selling conversation and slightly delays partnership terms.
A — Paid development per milestone watching
Immediate revenue, no upside beyond reputation and reusable components; scope-creep risk on hard 2D→3D work.
B — Partnership (equity / revenue share) watching
Uncapped upside if the product thesis holds; deferred, uncertain revenue and IP/decision-rights risk. Only with written IP terms, vesting and a defined exit.
C — Hybrid: paid build + small equity kicker (DEC-001 alternative) watching
Cash now plus optionality; adds term-sheet complexity to a relationship not yet scoped.
Pure-ML 2D→3D reconstruction (DEC-002) rejected
Scored 2.4 vs hybrid 4.3 (provisional); research-grade effort unless the documents reveal a constrained drawing family.
🚦 Decision Rules (pre-committed)
✅ Proceed when
Docs + structure + funded spike agreed — then treat as active build bet
⏸ Park when
No docs/structure agreement by 2026-10-03 — park without guilt; the idea keeps
⛔ Kill when
Docs reveal research-grade ambition with no funding, or IP terms unacceptable
💡 Advisor note
2026-09-12: Fastest unblocking sequence in the whole portfolio: one export session + one conversation. Until both happen, spend zero build hours. Authoritative registers: projects/Mario Software/03_Registers/.
8 actualisation steps · 2 done · 0 active · next: Hold the 60-minute scope conversation (ACT-002)
🧭 Dependency Map — Steps to Actualisation
Arrows show what each step needs before it can start; the longest left-to-right chain is the critical path.
📅 Stages & Time
🪜 Steps to Actualisation
Click any step to expand its sub-steps, owner and dependencies.
Export both Google Docs into 04_Inputs (ACT-001) done
- Done 2026-09-15 via the signed-in browser session; links + how-to in GDOC_LINKS.md
- Local Drive placeholders kept double spaces in filenames — that was why direct reads failed before export
Re-assess fully against the actual documents done
- Extract target user, domain, drawing types, pipeline ambition (automatic vs AI-assisted)
- Kill trigger: research-grade ambition with no funding
- Done 2026-09-15: §8 re-assessment appended — docs specify FULL automatic 2D→3D (research-grade, zero demand evidence); Mario's own PoC slice is the tractable path
Hold the 60-minute scope conversation (ACT-002) planned
- Cover Mario's role, funding, time commitment and expectations of Zaid
Decide and write down the deal structure (ACT-003 / DEC-001) planned
- Paid vs partnership vs hybrid; IP terms and walk-away rights if partnership
- Every hour before agreement is unpriced — hard cap 20 h
Define the MVP slice: one drawing family → one automated report (ACT-004) planned
- Full automatic 2D→3D is research-grade; a narrow slice is buildable and sellable
Gate review: Concept → Validation (due 2026-10-03) planned
- Docs read and re-assessment updated; conversation held; structure in writing; slice defined
- Miss → park without guilt; the idea keeps, the hours don't
Sell the paid feasibility spike (ACT-005, by 2026-10-31) planned
- Converts an open-ended favour into structured paid work — the first priced revenue
- Due 2026-10-31, four weeks after the gate it follows — date conflict tracked as ISS-03
Deliver the spike: technical approach + effort estimate planned
- Quote the real engagement per milestone against the written slice
- If auto-3D needs research-scale effort → reduce scope or walk
🔁 Development Process
Evidence-before-build, stage-gated: nothing is built until the documents are read, the deal is written and one narrow slice is chosen. Provisional technical approach (DEC-002): parametric/rule-based core with AI-assist where reliable — not full ML reconstruction. Any real engagement is quoted per milestone against a written slice, never time-and-materials against an open-ended framework.
Concept assessment
Provisional advisor assessment from titles/metadata; registers seeded; document links recovered and export how-to written.
- provisional ADVISOR_ASSESSMENT.md
- ACT/RSK/DEC registers
- GDOC_LINKS.md export how-to
Inputs & structure
Export and read both documents; full re-assessment; scope conversation; written structure; MVP slice defined.
- full re-assessment
- written deal structure
- MVP slice definition
Feasibility spike
Paid 5–20 h on one drawing family → one report: technical approach + effort estimate.
- technical approach
- effort estimate
- pricing evidence
Validation → Build (conditional)
Only if the spike sells and effort fits Mario's funding: build the slice, then get it used on real drawings by someone other than Mario.
- working MVP slice
- first non-Mario usage
Two principals and a dev machine — Zaid (capped hours), Mario (commitment unknown until the scope conversation) — plus fortnightly AI-advisor review. No cash tooling.
Two principals and a dev machine — Zaid (capped hours), Mario (commitment unknown until the scope conversation) — plus fortnightly AI-advisor review. No cash tooling.
💵 Budget
Total budget: $0 · Source: Advisor assessment §3 (2026-09-12): cash ~$0 unless hosting/models enter the spike
Allocation by phase (bar = allocated, amber = spent)
The true currency is founder hours — 1 of the 20-hour unstructured cap used. Any hosting or model-API cost is quantified at spike scoping, not before.
🧾 Investment Ledger
Capital at risk: Time only. Hard cap: 20 unstructured hours before deal structure is agreed in writing.
👥 People & Tools
People
| Name | Role | Commitment | h/week |
|---|---|---|---|
| Zaid Al Dabbag | Founder / AI-automation build capability | 1–2 h/wk while blocked (portfolio priority 2) | 2 |
| Mario | Counterparty — methodology & domain owner | Unknown — funding and hours to establish in the scope conversation (ACT-002) | 0 |
| AI advisor (ZCode) | Advisor / portfolio review | Fortnightly flip-through | 1 |
Tools
| Tool | Purpose | Cost |
|---|---|---|
| Mario's two framework documents (Google Docs) | The source specification — everything depends on them | free; exported 2026-09-15 |
| Existing dev machine + AI tooling | Spike delivery environment if sold | ~$0 unless hosting/models needed |
| Project registers (ACT/RSK/DEC) | Authoritative tracking | none |
🤝 Stakeholders
Keep informed · high interest / low influence
AI advisor (ZCode) championManage closely · high / high
Zaid Al Dabbag championMario supportiveMonitor · low / low
Prospective end users (unknown) neutralKeep satisfied · low interest / high influence
none mappedFull stakeholder register (4)
| Name | Role | Interest | Influence | Needs / notes |
|---|---|---|---|---|
| Zaid Al Dabbag | Founder / technical partner candidate | high | high | Evidence before hours; written structure; capped exposure |
| Mario | Counterparty — methodology author, domain expert | high | high | Scope conversation; honest feasibility feedback; a clear structure proposal |
| Prospective end users (unknown) | Future buyers/users of the tool | med | med | Unknown — documents (read 2026-09-15) name no end user or demand evidence (§8) |
| AI advisor (ZCode) | Advisor / portfolio review | high | low | Docs exported before re-assessment; spike-or-nothing discipline |
6 risks (3 high/critical) · 4 upside opportunities · 4 assumptions · 3 open issues
⚠️ Risk Register
L = likelihood, I = impact (1–5). Tint = severity band: low / medium / high (L×I ≥ 12) / critical (≥ 20).
Full risk register (6)
| ID | Risk | Category | L/I | Severity | Mitigation | Owner | Status |
|---|---|---|---|---|---|---|---|
RSK-01 | Technical feasibility overpromised — full automatic 2D→3D reconstruction is research-grade. | technical | 4/4 | high | Narrow MVP slice (ACT-004); parametric/rule-based core with AI-assist where reliable; paid feasibility spike before outcome commitments. | Zaid | open |
RSK-02 | Deal structure left undefined while work begins — unpriced hours accumulate silently. | commercial | 4/3 | high | Hard cap of 20 unstructured hours; written agreement gate (ACT-003); spike is paid regardless of structure. | Zaid | open |
RSK-03 | IP ownership ambiguity over the code Zaid contributes. | legal | 3/4 | high | Written IP terms before Build stage; vesting for time contributions if partnership; define walk-away rights. | Zaid | open |
RSK-04 | Market unknown — established BIM/CAD vendors occupy adjacent space; no customer evidence yet. | market | 3/3 | med | Extract the target user from the documents (ACT-001); name 3 prospective users before Build. | Zaid | open |
RSK-05 | Key-person dependency on Mario — his domain knowledge is both the moat and the single point of failure. | operational | 3/3 | med | Codify his process mapping into the product from day one; document decisions. | Zaid | open |
RSK-06 | Payment risk if Mario is bootstrapping — deferred or missed payments on a partner-level commitment. | commercial | 2/4 | med | Milestone-based invoicing; spike payment up front; structure decision includes funding reality. | Zaid | open |
🌱 Upside Opportunities
Cheapest unblock in the portfolio
One export session + one conversation produce a clear commit/no-commit — the fastest clarity per hour available anywhere in the portfolio (advisor note, 2026-09-12).
Upside: med · Effort: low
Paid feasibility spike
Converts an open-ended favour into structured paid work (5–20 h) that prices the real engagement regardless of which structure follows.
Upside: med · Effort: low
Reusable 2D→3D pipeline components
Under any structure, the engineering accrues to Zaid's AI-build capability and reputation (Option A upside in the advisor assessment).
Upside: med · Effort: med
Partnership upside if the thesis validates
Uncapped only with written IP terms, vesting for time contributions, a defined exit and a validated narrow MVP — the conditions are the strategy.
Upside: high · Effort: high
🧩 Assumptions
🔴 Issue Log (live)
| ID | Issue | Impact | Raised | Status | Resolution |
|---|---|---|---|---|---|
ISS-01 | Both framework documents are private Google Docs — unreadable from this machine until exported (ACT-001). | Blocks scope, pricing, structure, feasibility and the full re-assessment — the whole evidence base | 2026-09-12 | resolved | Resolved 2026-09-15: both docs exported via signed-in browser session and read; re-assessment §8 appended. |
ISS-02 | Deal structure undecided (DEC-001 parked) while the 20-hour cap clock runs. | Every contributed hour is unpriced; the gate cannot pass without a written structure | 2026-09-12 | open | |
ISS-03 | Register dates conflict with the Concept→Validation gate (due 2026-10-03): ACT-005 'sell the paid spike' is due 2026-10-31, and the MVP-slice due date differs between the KPI/register (ACT-004, 2026-10-10) and the gate (2026-10-03). | A 2026-10-03 gate review cannot find the spike sold or, on the register's own dates, the slice defined; either the gate defers those criteria or the register dates pull in | 2026-09-13 | open |
Stage KPIs · next gate: Concept → Validation (due 2026-10-03) · 4 open owner actions · full progress log.
📊 Stage KPIs
| KPI | Current | Target |
|---|---|---|
| Source documents obtained | 2 / 2 | 2 by 2026-09-26 |
| Deal structure agreed in writing | No | Yes by 2026-10-03 |
| MVP slice defined | No | 1 slice by 2026-10-10 |
| Unstructured hours burned | 1 / 20 | stay under cap |
🚪 Next Gate — Concept → Validation
Due: in 3d (2026-10-03)
- Both documents read and re-assessment updated
- Scope conversation held
- Deal structure agreed in writing
- MVP slice defined (one drawing family → one report)
✅ Next Steps & Actions
| ID | Action | Due | Status |
|---|---|---|---|
ACT-001 | Export Mario's two Google Docs into 04_Inputs (links in GDOC_LINKS.md) | overdue 4d (2026-09-26) | done |
ACT-002 | 60-minute scope conversation with Mario (user, drawing types, funding, expectations) | in 3d (2026-10-03) | open |
ACT-003 | Decide + write down deal structure: paid vs partnership | in 3d (2026-10-03) | open |
ACT-005 | Sell the paid feasibility spike | 2026-10-31 | open |
ACT-006 | Define the MVP slice with Mario — candidate from his own roadmap: single-storey floor plan → basic 3D → BoQ extract (ACT-004 in project register) | 2026-10-10 | open |