TBD Rubric - Bliss Rough Draft

Project TBD

RDI Decision Matrix — Criterion Definitions

J Bliss Perry + Dustin J Ross | Stanford GSB | STRAMGT 546 | Spring 2026

How to Use

This document defines the scoring criteria for Project TBD’s RDI Decision Matrix. Each criterion is rated 1–5, where 5 is most favorable. A score of 3 can also mean “unknown / more research needed.” A score below 3 in any criterion is disqualifying. For each criterion, concrete definitions are provided for scores 1, 3, and 5, calibrated to Project TBD’s specific context.

Score independently first, then compare. Where scores diverge by 2+, discuss — the conversation is as valuable as the number.

MARKET & TIMING

Semiconductor Supply Chain TAM

5	TAM exceeds $5B across compliance + intelligence + financial products; clear expansion driven by CHIPS Act, EU Battery Passport, UFLPA enforcement, and allied reshoring programs.
3	TAM is $1–5B and growing moderately; compliance market is real but hard to size precisely.
1	Addressable market is niche (<$500M) with flat or declining trajectory.

Problem Urgency / Acuteness

5	Firms are actively failing compliance audits, facing shipment holds at customs, or scrambling after supply shocks (e.g., Hualien earthquake) with no adequate tooling — the pain is acute today.
3	Problem is recognized and growing; regulatory deadlines are approaching but not yet binding for most firms.
1	Pain exists but is low-priority; companies manage with existing tools for the foreseeable future.

Regulatory & Geopolitical Tailwinds

5	Multiple reinforcing tailwinds — CHIPS Act funding conditions, UFLPA enforcement expansion, EU Battery Passport timelines, CFIUS scrutiny increases, allied export controls — creating compounding compliance demand over the next 5+ years.
3	Some tailwinds exist (e.g., one major regulation) but the trajectory could stall or reverse.
1	Regulatory environment is stable or deregulating; no new mandates expected.

PROBLEM QUALITY

Problem Prevalence Across Target Customers

5	Every firm in the semiconductor value chain — fabs, fabless designers, distributors, OEMs, contract manufacturers — must perform these compliance workflows; the problem is structurally universal.
3	The problem affects a significant portion of semiconductor firms but may be limited to certain tiers or geographies.
1	Only a narrow slice of semiconductor firms face this problem (e.g., one sub-segment).

Demand for a New Entrant vs. Incumbent Extension

5	No existing company has the right data architecture, AI-native design, or compliance-to-intelligence wedge to solve this; the opportunity structurally demands a new entrant.
3	Incumbents could pivot but haven’t prioritized it; a new entrant with a differentiated approach has a window.
1	Incumbents (SAP, Oracle, Resilinc) are well-positioned and already iterating toward this solution.

Problem Is Unavoidable / Mandated

5	Non-compliance carries legal penalties, shipment seizures, loss of CHIPS Act funding, or debarment from government contracts — the problem is mandated by law and unavoidable.
3	There are regulatory incentives but enforcement is inconsistent; customers can delay.
1	Customers can choose to ignore the problem or solve it informally.

DEFENSIBILITY & MOAT

Compounding Proprietary Data Asset

5	Every compliance check, every supplier onboarded, every shipment verified adds a unique node/edge to the ontology — the dataset becomes exponentially harder to replicate with each customer, creating a self-reinforcing data moat.
3	Some proprietary data is generated but it’s unclear how quickly it compounds or how defensible the dataset becomes.
1	Data generated is commoditized or easily replicated; no accumulation advantage.

Network Effects (Multi-Tenant Platform)

5	Cross-customer intelligence is core to the product — disruption alerts improve for everyone as the network grows; the recommendation marketplace only works because multiple customers share the ontology; each new customer makes the platform more valuable for all existing customers.
3	Some shared benefit exists (e.g., better benchmarks) but network effects are indirect or weak.
1	Each customer’s data is siloed; no cross-customer value creation.

Wedge-to-Platform Progression

5	The compliance wedge naturally and inevitably produces the data that powers every subsequent product layer (alerting → prediction → remediation → marketplace → financial products); the progression is sequenced and each step unlocks the next.
3	There’s a plausible roadmap from the wedge to a broader platform, but significant product leaps are required.
1	The initial product is a point solution with no clear path to something larger.

PRODUCT & TECHNOLOGY

Technical Feasibility with Current Technology

5	LLMs for document extraction, knowledge graphs for ontology construction, and agentic AI frameworks all exist today; the challenge is execution and integration, not invention.
3	The solution is technically possible but requires significant integration work and novel data pipelines; some components are unproven at scale.
1	Core technology doesn’t exist yet or requires fundamental research breakthroughs.

AI-Native Differentiation

5	The product is designed from the ground up for AI interaction — agentic simulation, conversational scenario modeling, LLM-readable ontology structures — and could not exist in a pre-AI world. AI is the product, not a feature.
3	AI meaningfully improves the product but the core value proposition works without it.
1	AI is a feature add-on to an otherwise traditional workflow tool; could be built without AI.

MVP-to-Full-Platform Path

5	The compliance MVP (supplier onboarding + export control verification) immediately generates production data that feeds the digital twin; every early customer interaction builds toward the full intelligence platform with no wasted motion.
3	The MVP demonstrates part of the vision but significant engineering and go-to-market work separates it from the full platform.
1	The MVP is disconnected from the long-term vision; major pivots needed to reach the platform.

BUSINESS MODEL & REVENUE

Path to Meaningful Early Revenue

5	Compliance SaaS generates revenue from Day 1 of customer deployment — firms will pay to automate UFLPA screening, ITAR/EAR verification, and mandatory disclosure filing because the alternative is manual labor and legal risk.
3	Revenue is achievable within 18–24 months but unit economics are uncertain; depends on pricing assumptions.
1	No clear monetization before the full platform is built; requires years of R&D before first dollar.

Venture-Scale Outcome Potential

5	The compliance-to-intelligence-to-financial-products progression creates multiple $1B+ revenue opportunities; the digital twin as financial infrastructure is a category-defining outcome.
3	Could become a solid business ($100M–$500M revenue) but the path to a $1B+ outcome requires significant market expansion.
1	The market ceiling is too low to support a venture-scale outcome (<$100M revenue potential).

Multiple Monetization Layers

5	Clear progression from compliance SaaS → intelligence platform subscriptions → marketplace transaction fees → financial product origination fees — each layer higher margin and more defensible than the last.
3	Two revenue streams are credible (e.g., compliance SaaS + intelligence subscriptions).
1	Single revenue stream (e.g., compliance SaaS only) with no clear expansion.

VENTURE FUNDABILITY

Exit Magnitude & Legibility

5	The company could be a standalone public company (Palantir-scale defense/intelligence platform) or command a transformative acquisition price; the ‘Bloomberg Terminal for supply chains’ narrative is immediately legible to growth investors.
3	M&A exits are plausible (Palantir, Bloomberg, Resilinc acquire) but the public-company path is uncertain.
1	The exit story is unclear; hard to identify strategic acquirers or public-market comparables.

Milestone-Based Funding Stages

5	Clear pre-seed (demo + first pilot) → seed (10 paying compliance customers, digital twin prototype) → Series A (intelligence platform launched, government partnership LOI) → Series B (marketplace live, first financial product) progression with measurable milestones at each stage.
3	Funding stages exist but some rounds require leaps of faith about unproven product-market assumptions.
1	The capital requirements are front-loaded with no clear milestones to de-risk between rounds.

FOUNDER–MARKET FIT

Relevant Experience & Network

5	Founders bring GSB strategy + finance training, AI proficiency, Palantir network connections, Prof. Berk faculty sponsorship for downstream supply chain research, and access to Stanford’s semiconductor/policy ecosystem.
3	Founders have adjacent experience (finance, technology) and some network connections, but lack deep domain expertise in semiconductor supply chains.
1	Founders have no background in semiconductors, compliance, or enterprise SaaS; would be starting from zero.

10-Year Passion & Energy

5	Founders describe this as ‘building the financial nervous system of the global economy’ and articulate a vision of American industrial leadership — this is a mission, not a project. The ‘proud parent test’ passes clearly.
3	Founders are intellectually interested but haven’t tested whether this specific problem sustains their energy through hard years.
1	The opportunity feels like a means to an end; founders would jump to something else if a better option appeared.

Technical + Industry Expertise Access

5	The team can build AI/data products directly; has identified specific domain experts to recruit (Holland for CS, Danny Huckins for customs/trade, Jenny Steiger for AI/GSB); and the independent study with Prof. Berk provides structured access to industry insiders.
3	The team has some capabilities but critical gaps remain (e.g., deep semiconductor manufacturing knowledge, compliance regulatory expertise).
1	The team lacks both the technical skills and industry knowledge needed, with no clear path to acquiring them.

STRATEGIC OPTIONALITY

Adjacent High-Upside Opportunities

5	The digital twin / ontology unlocks at least four distinct high-upside paths — parametric insurance, prediction market middleware, government design partnerships, and capacity intermediation — each independently venture-scale, and the core compliance data naturally produces the inputs needed to pursue them.
3	One or two adjacencies exist but they require significant new capabilities or partnerships.
1	The core product is self-contained; no clear ‘big swing’ adjacencies.

Wedge Produces Future Optionality

5	The compliance wedge is specifically designed to generate the structured data, customer relationships, and regulatory trust needed to pursue financial products, government contracts, and marketplace intermediation — optionality is engineered into the strategy, not accidental.
3	The wedge creates some strategic positioning but the leap to big swings requires deliberate pivots.
1	The initial product generates revenue but doesn’t position the company for anything larger.

GO-TO-MARKET CLARITY

Clear First Customer & Repeatable Path

5	First customers are semiconductor firms with acute compliance pain (UFLPA screening, ITAR verification); the RDI interview program and Prof. Berk’s network provide direct access; each landed customer generates a referenceable case study and data for the next sale.
3	First customer type is identified (e.g., mid-size semiconductor distributors) but the path to the next 10 customers is unclear.
1	No identified first customer segment; GTM is theoretical.

Dual-Demo Sales Strategy

5	The agentic simulation demo (synthetic data, scenario modeling) excites executives and investors about the destination; the compliance workflow demo (supplier screening, export control automation) sells the practical first contract — together they tell a complete story.
3	Can articulate both the vision and the practical product, but the demos feel disconnected.
1	Can only sell the long-term vision OR the practical first step, but not both.

Revenue Before Full Platform

5	The compliance SaaS product generates standalone recurring revenue from Day 1 — supplier onboarding and ITAR/EAR screening are valuable products even without the full intelligence platform, and every dollar of compliance revenue simultaneously builds the digital twin.
3	Some early revenue is possible but it requires significant customization or one-off engagements.
1	The GTM requires building the entire platform before any revenue is possible.

RISK & RED FLAGS

Regulatory Risk (Policy Reversal)

5	The thesis is supported by multiple independent regulatory frameworks across multiple jurisdictions (CHIPS Act, UFLPA, EU Battery Passport, CFIUS, ITAR/EAR) — no single policy reversal undermines the opportunity, and the trend toward more regulation spans both parties and allied governments.
3	The thesis benefits from current regulations but could survive partial rollbacks; some diversification across regulatory regimes.
1	The thesis depends on a single regulation that could be repealed or weakened; deregulation would be fatal.

Concentration Risk (Geography/Customer)

5	Multi-jurisdictional from inception (U.S. + EU + allied nations all implementing parallel compliance regimes); customer base spans fabs, fabless, distributors, OEMs, and government agencies — no single concentration point can kill the business.
3	Moderate concentration — primarily U.S.-focused initially, but with a credible international expansion path.
1	Dependent on a single geography, single customer type, or single regulatory regime.

Execution Complexity

5	The sequenced approach (compliance MVP → data accumulation → intelligence products → financial infrastructure) means only one thing needs to work at each stage; each milestone de-risks the next, and early revenue from compliance SaaS funds continued development.
3	Execution is challenging but the sequenced roadmap reduces the number of parallel bets; some critical dependencies remain.
1	Requires simultaneous breakthroughs in technology, regulation, sales, and product — too many things must go right at once.

Competitive Risk (Incumbent Pivot)

5	The AI-native architecture, compliance-as-data-acquisition model, and multi-tenant ontology represent a fundamentally different approach than incumbents’ architectures; by the time incumbents recognize the threat, the compounding data moat makes catch-up extremely difficult.
3	Incumbents could compete on parts of the product (e.g., compliance automation) but would struggle to replicate the full wedge-to-platform data strategy.
1	A well-funded incumbent (SAP, Resilinc, Palantir) could replicate this approach within 12–18 months.

Geopolitical Assumption Dependency

5	The thesis is built on structural regulatory trends (compliance mandates, reshoring incentives, export controls) that exist independent of any single crisis — the product is valuable in peace and indispensable in crisis.
3	Geopolitical tailwinds strengthen the thesis but the core compliance product has value regardless; some scenarios could reduce urgency.
1	The entire thesis collapses if Taiwan Strait tensions ease or a single geopolitical scenario doesn’t materialize.

QUALITATIVE GUT CHECK

What excites you most about this idea? What keeps you up at night?

Would you be proud to describe this to someone you respect? (The ‘proud parent test’)

What is the single biggest unknown that research could resolve?

If this works, what does the company look like in 10 years?