DomesticProof — The BABA Domestic-Content Qualification & Certification Engine

02Final decision: Blueprint

This run produces a go decision. The candidate cleared the evidence threshold: a named buyer with revenue at stake, a fresh statutory trigger with a hard date (October 1, 2026), verified enforcement fear, verified incumbent budget (national CPA-firm advisory practices, a funded AI software startup, and internal compliance labor), a narrow one-artifact MVP, and a credible path to 50%+ gross margin with expert judgment concentrated at one signing chokepoint.

03Executive summary

04Thesis

05Discovery rationale

This run searched six sectors before selecting: dental revenue-cycle operations, college-athletics NIL compliance, auto-finance GAP refund remediation, local-government public-records processing, Medicare TPMO marketing oversight, and federal domestic-preference compliance. Fourteen targeted searches were performed (see research ledger). BABA qualification won because it uniquely combines: (a) a dated statutory trigger 12 weeks away creating urgency for a population that has never been regulated before [S1]; (b) revenue capture, not just risk avoidance — no certification letter means lost sales today [S22]; (c) a pure self-attestation regime with no licensed-professional gate, so a documentation service faces no unauthorized-practice barrier [S8]; (d) verified budget via CPA-firm advisory practices and a venture-funded software category [S12][S13]; and (e) zero overlap with the 187 prior blueprints in the manifest, which cover import-side trade (UFLPA, AD/CVD, HTS, first-sale, FTZ, duty drawback) but nothing on domestic-content certification for federally assisted projects.

06Candidate comparison

Losers were rejected on evidence, not preference: each tripped at least one of shrinking-pain, incumbent-density, government-cycle, or whitespace criteria that BABA cleared.

07CODE validation

08Rubric scorecard (six gates + factory rubric)

09Target buyer

10Jobs-to-be-Done

• "Tell me, product by product, whether I can legally sign this certification letter — and give me the file that proves it."
• "Get my products qualified before October 1, 2026 so my distributors stop dropping me from federally funded bids."
• "Chase my 40 suppliers for origin attestations so my team doesn't have to."
• "Keep my certifications current when my BOM, suppliers or the rules change — and re-issue letters automatically."
• "Give my contractors/grantees exactly the letter format their agency requires (EPA vs FHWA vs state CDBG) without me learning five regimes." [S9][S15]
• "If an auditor or a qui tam relator ever asks, hand me the defensible methodology memo." [S6][S16]

11Painful problem

12The outcome we sell

13First one-feature MVP wedge

14Evidence summary

• Verified FHWA final rule with phased 10/1/2025 and 10/1/2026 requirements [S1][S3][S17]; 1983 waiver history [S2].
• Verified $711.8B BABA-covered IIJA grant pool; $131.2B newly available FY2026; 68k+ projects [S4][S5].
• Verified Self-attestation regime, agency cert-letter templates, audit retention by grantees [S8][S9][S15].
• Verified Enforcement fear: record $6.8B FCA year, 1,297 qui tams, criminal indictment for false domestic-sourcing certs, $14,308–$28,619 per-claim penalties [S6][S7][S19].
• Verified Budget exists: national CPA advisory practice [S12]; Certivo $4M seed SaaS [S13]; AGC member resource burden [S14].
• Inferred Category spend $300M–$1B+/yr; thousands of newly covered manufacturers; mid-market lacks component-level cost tracing.
• Unverified (kill tests): scan→paid conversion ≥25%; $2,500–$7,500 per-family price acceptance; reviewer minutes 240–420→90–150; renewal ≥70%.

15Claim table

16Source-claim matrix

17Market and demand evidence

Demand is visible on both sides of the letter. On the requesting side, agencies and states publish certification templates and instruct grantees to collect and retain them [S9][S15]; AGC runs a member resource hub because contractors must chase these documents on every funded job [S14]. On the supplying side, Commerce and NTIA maintain public self-certification lists that function as de-facto "approved vendor" directories — being absent is a competitive handicap [S10][S11]. The consulting and software categories already monetize the pain [S12][S13]. What does not exist at scale is the mid-market, fixed-price, done-for-you qualification desk.

18Active buyer conversations

• AGC's BABAA hub and state AGC chapters field continuous member questions on cert collection and product coverage [S14].
• EPA felt compelled to publish a manufacturers' FAQ (June 2025) — direct evidence manufacturers are asking agencies how to certify [S8].
• Law-firm alert volume (Crowell, Wiley, Ballard Spahr, Pillsbury, K&L Gates, Holland & Knight) around the FHWA rule and FCA risk signals paying-client demand for exactly this question [S2][S19][S20].
• State program offices (Indiana OCRA, Minnesota NAHRO, NC DEQ) publish contractor-facing BABA forms because submissions keep failing at pay-application time [S15].
• Certivo's seed round and content engine targeting "Buy America compliance in 2026" queries shows commercial search demand worth investing against [S13][S21].

19Competitive landscape

20Competitor and budget validation

Existing budget sources: (1) advisory fees already paid to CPA firms for BABA assessments [S12]; (2) SaaS subscriptions for supplier-attestation software [S13]; (3) internal labor — quality managers and sales ops chasing supplier letters by email; (4) legal spend when a certification goes wrong [S7]. Why current alternatives fail: consulting is episodic and expensive; software transfers the work and the liability back to the buyer; internal staff lack methodology and it shows at audit [S16]. Why we win: fixed price per product family, 10-day turnaround, an expert-signed determination the buyer can stand behind, supplier-chasing done for them, and a renewal subscription that keeps the file alive as rules escalate [S21]. We are not "no competitors" — we are a better-shaped vehicle for budget that verifiably exists.

21Pricing evidence and proposed pricing

22Regulatory and compliance considerations

• No licensing gate to prepare documentation. The BABA regime is manufacturer self-attestation [S8]; preparing cost analyses and draft letters is unregulated professional services work. The manufacturer's own officer signs outward certifications.
• Unauthorized practice of law: interpreting ambiguous regulatory text for a client edges toward legal advice. Mitigation: determinations apply published tests and agency guidance with citations; novel interpretation questions (e.g., "component vs. material" edge cases without guidance) trigger referral to government-contracts counsel; engagement letters state we do not provide legal advice.
• FCA accessory exposure: if we knowingly paper over bad data we could face liability. Mitigation: evidence-or-it-doesn't-count policy, client data attestations, documented refusal to issue when evidence is insufficient, E&O insurance, right-to-withdraw clause.
• Flat-fee integrity: no success fees tied to winning bids or passing audits — avoids incentive-to-shade findings and keeps the work product credible.
• Confidentiality: BOMs and component costs are crown-jewel data; NDAs, client-segregated storage, no cross-client data reuse without anonymized aggregation consent; SOC 2 roadmap.

23Licensing boundary

24AI-native advantage

AI changes the economics, not just the speed. A qualification file requires reading a 200-line BOM export, three agencies' rules, a dozen supplier PDFs in inconsistent formats, and a cost ledger — then producing math, citations and prose. Done by consultants this is 20–40 billable hours; done by our engine it is minutes of compute plus a senior review. AI tasks: BOM normalization, component classification, origin-evidence parsing, content-percentage computation drafts, rules retrieval, memo/letter drafting, supplier-chase emails, change monitoring. Human tasks: judgment calls at classification boundaries, methodology approval, determination signing, client relationship. Deterministic tasks: the percentage math itself, template validation, checklist gates, citation-link resolution. Data inputs: BOM + costs, supplier attestations, agency rules corpus. Output artifacts: workbook, determination memo, evidence index, letter set, gap plan. Failure risks: hallucinated citations (blocked by resolution checks), misclassified components (blocked by expert gate), stale rules (blocked by versioned rules corpus). Never fully automated: the determination sign-off and any "close call" classification.

25Internal AI engine architecture

26AI-vs-human operations pipeline

27Dynasty translation layer

28Anti-duplication analysis

Vs. the manifest (187 runs): first domestic-content / Buy America / BABA entry. Nearest neighbors and why they differ: uflpa-forced-labor-traceability (import ban evidence, CBP detentions — different law, buyer motion is imports); ad-cvd-importer-compliance (duty exposure); hts-classification (tariff codes); first-sale-export-valuation (customs valuation); sba-setaside-certification (contractor status certification, not product content); gsa-mas-contract (federal sales channel); dcaa-incurred-cost (cost accounting for contractors, not products). None sells a product-level domestic-content determination to manufacturers serving federally assisted projects.

Vs. the market: not a generic compliance dashboard (we deliver signed files, not software); not consulting (fixed-price, productized, 10-day SLA); not Certivo (they sell the buyer a tool and the buyer keeps the judgment burden [S13]). The under-served segment is the $10–250M manufacturer with no compliance staff. The unsolved pain is the defensible determination plus done-for-you supplier evidence. Our differentiation compounds through the per-agency acceptance playbooks and the classification precedent library no tool ships with.

29Anti-commoditization analysis

If future general models let a quality manager ask a chatbot "does my product pass the 55% test?", what survives? (1) Accountability: a chatbot answer is not a signed determination an officer will stake a federal certification on; (2) evidence logistics: models don't chase 40 suppliers for attestations, validate them, and keep them current; (3) the precedent library: thousands of reviewed classification calls and per-agency acceptance outcomes are proprietary training data for our engine, not the model's; (4) the subscription: thresholds escalate (55→65→75 [S21]) and BOMs drift — monitoring is a relationship, not a query; (5) the two-sided position: once distributors and grantees recognize our file format, "DomesticProof file attached" becomes the fast lane, a network effect no self-serve prompt replicates. Model improvement lowers our COGS faster than it improves the amateur's confidence to sign.

30Service delivery workflow

1. Intake (day 0–1): engagement letter + data request checklist; client uploads BOM/ERP export, cost ledger extract, supplier contacts, product specs, target programs. Deterministic completeness gate; gaps chased same-day.
2. Normalization (day 1–2): BOM to canonical schema; cost basis reconciled to ledger; anomalies flagged.
3. Rules mapping (day 2): program(s) → applicable tests (iron/steel 100% vs manufactured product 55% vs construction materials; FHWA vs EPA vs NTIA specifics [S1][S8][S10]).
4. Supplier evidence (day 2–7): AI-drafted, operator-sent origin attestation requests; replies parsed, validated, indexed; non-responders escalated to client with scripts.
5. Computation & classification (day 5–7): content percentages computed in code; boundary calls queued for the senior expert.
6. Determination & drafting (day 7–9): expert resolves calls, approves methodology, signs memo; letters drafted to each requesting format [S9][S15].
7. QA & red-team (day 9): checklist pass; a second reviewer attacks every "qualified" conclusion; citation resolver run.
8. Delivery & debrief (day 10): file pack + 30-min debrief; officer signs letters; gap plan walkthrough if not qualified.
9. Follow-up: 30-day acceptance check with requesting parties; monitoring subscription pitch; renewal calendar set.

31Operations as product

Variance is the enemy; the operation is the moat. Standing assets: intake checklists and required-evidence lists per program; SOPs per workflow stage; exception queues (classification, evidence, methodology); reviewer assignment by product vertical; confidence scoring on every AI proposal; full audit trail (who/what/when for every number in the memo); version-controlled rules corpus with effective dates [S21]; gold-standard example files per vertical; red-team checks on qualified conclusions; customer-ready templates per agency; root-cause analysis on any letter rejected by a requesting party; postmortem loop feeding SOPs, prompts and checklists. Target: every tenth file in a vertical is cheaper and more consistent than the first.

32No-holes quality engine

• Gate A — completeness: no computation until required inputs exist; missing cost data is a client-visible blocker, never an assumption.
• Gate B — math: percentages recomputed deterministically; AI draft must match to the cent or the unit fails.
• Gate C — evidence: every origin claim maps to a document in the index; "supplier says so verbally" does not pass.
• Gate D — citations: every rule reference resolves to the versioned corpus; dead or misquoted citations block release.
• Gate E — adversarial review: qualified conclusions get a second reviewer whose job is to break them.
• Gate F — signature discipline: no determination ships without the senior expert's sign-off recorded in the audit trail.

33What the human expert actually does

Operator (non-expert) minutes: 150 at launch → 60 at day 90 (chasing, spot-checks, packaging). Both tracks are day-90 kill tests (§49).

34Minimum viable offer

"BABA Fast-Track Qualification" — one product family, $2,500 founding rate (list $3,500), 10 business days, deliverables as in §13, plus a free Exposure Scan of the rest of the catalog to seed the next order. Cap: 10 founding accounts. Everything needed exists today: templates, rules corpus, AI workbench, one senior expert (founder or first hire), an e-sign tool and a spreadsheet.

35Fulfillment process (first 3 customers, manual-first)

1. Founder runs intake call personally; data request via checklist email.
2. BOM parsed with AI assist into the canonical workbook; classifications proposed by model, every one reviewed by the expert (100% review at launch).
3. Supplier chases sent from a branded mailbox with merge templates; tracked in the workbook.
4. Math in the workbook (formulas visible = client trust); memo and letters drafted by AI from the workbook, edited and signed.
5. Deliver by email + debrief call. Ask for the catalog order and a referral.

Don't automate yet: classification review (need the precedent data), debriefs, pricing. Automate later: BOM ingestion, chase sequencing, monitoring diffs, status reporting.

36Tools and systems

37Human-in-the-loop quality control

Launch posture: 100% expert review of all classifications and every memo. As the precedent library grows, review narrows to low-confidence and novel items — but three invariants never relax: (1) every determination is expert-signed; (2) every qualified conclusion survives adversarial second review; (3) every number traces to a document. Reviewer calibration: weekly gold-file exercises; drift measured by re-review sampling (5% of shipped files monthly).

38Nonlinear scaling and unit economics

Headline GM above is per-file contribution; blended company GM lands 55–65% after unsold capacity, scan fulfillment (free), sales support and monitoring-subscription service costs — still comfortably past the 50% bar. Automation share of pipeline steps: ~45% launch → ~70% day 90 → ~85% year 1. Cycle time 10 days → 5. Escalation-rate target <10%; quality-failure (letter rejected) target <5% with root-cause postmortems. CAC payback ≤2 files (outbound-led). Funnel assumptions (kill tests): scan→paid ≥25%; pilot→catalog ≥50%; monitoring attach ≥60%; renewal ≥70%. Margin expands with volume because the precedent library compresses expert minutes — the engine gets cheaper per unit as frontier models improve, satisfying the decoupling requirement: revenue grows with files and subscriptions, headcount grows with exception volume only.

39Distribution proof table

40Sales and outreach plan

Motion: Scan → Debrief → Fast-Track file → Catalog program → Monitoring. The free Exposure Scan does the qualifying: it names the products at risk, the tests that apply, and the missing evidence — pain made visible in the buyer's own catalog. The debrief call presents findings and quotes the founding rate on the highest-stakes family. Objection handling: "our lawyer can do it" → lawyers opine, they don't build evidence files, and we're 10x cheaper; "we'll wait" → projects obligated after Oct 1 are already being bid today; "we bought software" → who signs the determination and who chases your suppliers?

41Founder-led content plan

Position the founder as the translator of a confusing regime: weekly teardown posts (anonymized real BOMs walked through the 55% math), plain-English rule explainers per agency, "cost of doing nothing" pieces built on FCA cases [S7], misconception series ("final assembly ≠ compliant after Oct 1"), and buyer-question mailbags harvested from webinars. High-performing posts become paid-ad creative and webinar topics later.

42First 30 days of content

43Lead magnet and waitlist plan

Artifact before payment: the Exposure Scan memo — product-line risk table, applicable tests per program, missing-evidence list, and one worked example. Why it creates trust: it demonstrates the exact methodology the paid file uses, on their catalog. Pain signal captured: which families they ask about first = where revenue is at risk. Follow-up: debrief call within 5 days, founding-rate quote, 30/60-day nurture with rule-change alerts. Sales-ready when: a named product family + an active letter request or a bid pipeline into assisted projects + officer willing to sign. Waitlist mechanics: founding cohort capped at 10 accounts at $2,500; overflow joins a dated waitlist with a locked rate — scarcity is real (expert review capacity), not theater. Signups are treated as interest, not PMF; only paid conversion and renewal count (§53).

44Warm GTM plan

Week 1 outreach to: personal/professional network in construction products and gov contracting; webinar registrants; scan users; LinkedIn engagers; association staff contacts; and the referral-partner shortlist (3 gov-contracts attorneys, 3 regional CPA firms, 2 ERP consultancies). Offer: free scan + consultative debrief, founding rate for live files. Every scan debrief ends with two asks: the highest-risk family, and one distributor contact who's been demanding letters.

45Targeted outbound plan

List build: state DOT QPL/APL rosters × product verticals (ITS, lighting, drainage, precast accessories) filtered to $10–250M US manufacturers; NTIA/Commerce self-cert lists for who's conspicuously absent [S10][S11]; exhibitor lists from ATSSA/AWWA shows. Personalization: name the product, the list it appears on, and the specific test with its date. Lead with the diagnosis memo, never a demo ask. Sequence: email 1 diagnosis → email 2 worked example from their vertical → LinkedIn touch → email 3 founding-rate deadline. Volume: 30 accounts/week, quality over spray.

46Answer-engine / search visibility plan

Buyers now ask ChatGPT/Perplexity "does my product need to meet Buy America 55%?" — the sources those engines cite are law-firm alerts and agency PDFs [S2][S8]. Publish the clearest per-agency explainer pages on the open web (schema-marked, plainly written, with the actual thresholds and dates), a free calculator, and transparent pricing — the combination answer engines prefer to surface. Target queries: "BABA certification letter template," "FHWA 55% domestic content," "Buy America manufactured products 2026," "who certifies BABA compliance." Measure: assistant-referred scan signups (UTM + "how did you hear").

47Pilot design and early-demand trap mitigation

Founding cohort: 10 accounts max, one product family each, $2,500, 10-day SLA, in exchange for outcome data (was the letter accepted? by whom? any pushback?), a testimonial on success, and tolerance for a debrief questionnaire. The early-demand trap — mistaking deadline-panic signups for durable demand — is mitigated by instrumenting paid conversion, catalog expansion and monitoring attach as the real signals, and by capping the cohort so we learn instead of drowning. Custom work boundary: anything not producible from the standard file shape (e.g., waiver petitions, FCA response support) is referred out or scoped separately — pilots don't get bespoke engines.

48Early-access feedback flywheel

Weekly cohort review: every expert correction is classified (extraction error / classification call / methodology / drafting) and becomes a prompt fix, a rule, a checklist line, or a precedent entry — with an owner and a ship date. Client feedback channel: debrief questionnaire + 30-day acceptance check. Product feedback = anything about the standard file; custom work = new artifact requests (logged, not built). Fix-before-expand list: any Gate failure pattern, any letter rejection, any SLA miss ≥2 days.

49Build-before-scale checkpoints

• After 5 accounts: harden intake checklist, evidence minimums, BOM schema. Measure minutes per file honestly.
• After 10 accounts: harden SOPs, exception queues, reviewer checklist, letter templates per agency actually encountered. Kill tests due: scan→paid ≥25%? expert minutes ≤ launch targets? any letter rejected?
• After 20 accounts: pause new sales until COGS, rework, escalation and cycle-time are measured against §38. Acceptable temporary manual work: supplier chasing, packaging. Red-flag manual work (signals unscalable model): re-deriving methodology per client, per-client letter formats beyond the template set, expert re-reviewing high-confidence classifications.
• Repeated-fix rule: third occurrence of any correction type must ship as a system change, not a person remembering.

507-day launch plan

1. Day 1–2: finalize rules corpus v1 (2 CFR 184, M-24-02, FHWA rule, EPA/FEMA/NTIA guidance + templates); build canonical workbook + checklists.
2. Day 2–3: build Exposure Scan template + landing page with scan CTA; engagement letter + data-request pack; E&O quote started.
3. Day 3–4: build QPL-based outbound list v1 (100 accounts, 2 verticals); draft sequences; set up CRM + branded mailbox.
4. Day 5: publish 2 cornerstone explainers + the self-test worksheet; founder LinkedIn post 1.
5. Day 6–7: send first 30 outbound diagnoses; book association webinar slot; run 1 full dry-run file on a synthetic BOM through all gates.

5130-day launch plan

1. Weeks 2–3: deliver first 3–5 scans; convert ≥2 founding files; ship them through the full pipeline; collect acceptance outcomes.
2. Week 3: webinar #1 with association or partner; publish teardown #1; referral-partner kits to 8 firms.
3. Week 4: postmortem cohort #1; harden per §49-5; publish pricing page + per-agency guide #2; 120 cumulative outbound accounts; decide founding-rate extension based on funnel data.

5290-day launch plan

1. Reach 10 founding accounts; run 10-account checkpoint and kill tests (§49).
2. Launch Monitoring subscription to all qualified clients (attach target ≥60%).
3. Stand up supplier-chase automation + BOM ingestion scripts (biggest minute sinks).
4. Publish 6 per-agency/vertical guides + calculator; webinar cadence monthly; first distributor push-down partnership live.
5. Hire/contract second reviewer if expert utilization >70%; begin SOC 2 scoping.
6. Day-90 go/kill review against §53 gates; if green, open catalog programs at list pricing.

53Metrics and KPIs

54Risks and mitigations (summary)

The four that matter most: (1) willingness-to-pay is assumed, not proven — mitigated by the capped founding cohort and hard kill tests; (2) rule instability (EO-driven waiver changes, reauthorization politics [S18][S20]) — mitigated by the versioned rules corpus and monitoring subscription that monetizes change instead of suffering it; (3) data-quality dependence — client cost data may be garbage; mitigated by attestation requirements and refusal-to-issue discipline; (4) incumbent drift into our lane — Certivo adding services or CPA firms productizing; mitigated by speed, the mid-market price point, and the precedent library. Full register below.

55Exhaustive risk register

56What could kill this

Honestly: (1) the mid-market deciding a template letter signed blind is good enough until someone in their vertical gets sued — timing risk we mitigate but can't control; (2) a broad regulatory retreat restoring the manufactured-products waiver; (3) failure to compress expert minutes, leaving us a boutique consultancy with AI garnish; (4) a funded competitor giving away qualification as a loss-leader for software seats. Each has a named kill test or watch-trigger in §49/§53/§55.

57Go/no-go reasoning

Go. Evidence threshold met on every element: named buyer (VP Sales at newly covered manufacturers), specific painful problem (can't defensibly sign demanded certifications; 55% test dated), verified existence and spend (agency templates, CPA advisory, funded SaaS, FCA fear [S8][S12][S13][S7]), narrow MVP (one file, one family, 10 days), practical first-sale path (QPL outbound + scan), no fatal regulatory blocker (self-attestation regime; no license required to prepare documentation), credible 50%+ GM path (§38), believable distribution (§39). Honest weaknesses — unproven price acceptance, deadline-dependent urgency, expert-minute compression — are instrumented as kill tests rather than hand-waved.

58Final recommendation

59Source list

1. Federal Register — Buy America Requirements for Manufactured Products (FHWA final rule), Jan 14, 2025
2. Crowell & Moring — End of the Road: FHWA Rescinds Longstanding Buy America Waiver
3. TRUSS FABER — BABA: Key 2025 and 2026 Rule Changes
4. GAO-25-107243 — Infrastructure Grants: Status of Funding as of Dec 31, 2024
5. US DOT — IIJA Funding Status
6. Wiley — Key Takeaways from DOJ's FY 2025 False Claims Act Report
7. Mondaq — Manufacturers See Rising FCA Risk With U.S.-First Laws
8. EPA — BABA FAQs for Manufacturers (June 2025)
9. EPA — Manufactured Product Certification Letter Template
10. NTIA — BABA Compliance and Reporting Requirements (Oct 2025)
11. US Commerce — BABA Self-Certification List
12. Cherry Bekaert — BABA: Key Compliance Rules
13. BusinessWire — Certivo Raises $4M Seed Round (Feb 2026)
14. AGC — BABAA Resource Hub
15. Indiana OCRA — BABA Manufacturer Certification Letter / MN NAHRO — BABA Contractor Compliance Form
16. Cherry Bekaert — BABA Basics & FAQs
17. FHWA — Announcement: Updates to Buy America Requirements
18. ConstructionOwners — Infrastructure Funding Cliff 2026
19. K&L Gates — The Expanding False Claims Act (Jun 2026)
20. Holland & Knight — EO Increases Scrutiny on Buy American (Mar 2026)
21. Certivo — Buy America Compliance in 2026
22. FedBizAccess — BABA Explained for Small Businesses
23. eAssist — Dental Insurance Verification Pricing / Teero — Outsourcing Dental Insurance Verification
24. College Sports Commission — Revenue Sharing / Bradley — Enforcing After House
25. CFPB — Supervisory Highlights, Auto Finance Special Edition / Allied Solutions — The GAP Trap
26. CivicPlus — Current Issues in Public Records Request Management
27. eCFR — 42 CFR Part 422 Subpart V (MA Communication Requirements) / Ritter — Medicare Call Recording FAQs

Generated 2026-07-07 by the AI-Native Service Business Discovery & Blueprint Factory. Claims labeled Verified / Inferred / Unverified per the source-claim matrix (§16). This document is business research, not legal advice.