AI-native service blueprint · Hardened v2

BABA Domestic Content Compliance Engine (OriginVault)

Done-for-you Build America, Buy America (BABA) domestic-content documentation, manufacturer-certification chase, payment-package assembly, and waiver-packet prep for mid-market infrastructure contractors — not a customer-operated compliance co-pilot.

Run 2026-07-09T18:02:00Z Decision: Blueprint Slug: baba-domestic-content-compliance-engine Sector: Federal infrastructure / construction compliance Pricing: per-project package + per-SKU cert chase

01 Thesis

Mid-market general contractors and specialty subcontractors on federally assisted infrastructure projects already lose progress payments, burn project-controls hours, and carry False Claims Act exposure when manufacturer BABA certifications are missing, generic, mis-categorized, or late. OriginVault sells the payment-ready BABA evidence package — product-classified SKU register, project-specific manufacturer certifications, mill-test / component-cost support where required, exception queue, and optional waiver packet — assembled by an internal AI engine and signed off by a domestic-content specialist. The customer experiences a compliance desk, not software.

Why AI-native: the work is document-heavy extraction, category classification under 2 CFR Part 184 / agency overlays, supplier chase orchestration, completeness scoring, and audit-trail packaging — exactly the pattern frontier models accelerate while humans remain at categorization edge-cases, waiver strategy, and customer-facing trust. Why not a co-pilot: primes and owners already withhold payment until the file is complete; they will not staff another portal.

02 Discovery rationale

Terrain scanned this hour: BABA/domestic content, EPCRA Tier II (already crowded with consultants), AES/EEI export filing (commoditized forwarder work), FTC Click-to-Cancel (federal rule vacated — softener), Section 45Q (narrow taxpayer set + licensed tax practice), LIHTC/HUD (prior HOTMA/LIHTC blueprints), OSHA ITA (prior), PFAS TSCA (prior + delayed start), CA pay data (H17 OFF-LIMITS).

  • Decisive verified catalyst: FHWA Manufactured Products Final Rule (effective Mar 20, 2025) phases in US final-assembly for manufactured products on obligations on/after Oct 1, 2025 and the 55% domestic component-cost test on/after Oct 1, 2026 — expanding documentation burden beyond traditional iron/steel. Verified
  • Budget proof: Big-4/advisory (Cherry Bekaert), sourcing consultants (Free Market America), and manufacturer-letter marketplaces already sell BABA help — category spend exists; mid-market DFY whitespace remains. Verified
  • Payment pain: State CDBG / municipal specs explicitly withhold progress payments until BABA certifications are received. Verified
  • Anti-dup: Distinct from IRA-PWA (payroll/apprenticeship for tax credits) and prevailing-wage certified payroll — different statute (BABA §§70901–27 / 2 CFR 184), inputs (manufacturer certs, mill tests, component-cost), buyer job (unlock payment / survive audit), adversary (agency payment hold + FCA on false domestic-origin claims).

03 Candidate comparison

CandidateBuyerOutcomeScore /5 avgEvidenceDecision
BABA Domestic Content Desk (OriginVault)Mid-market GC / specialty sub on FHWA, EPA SRF, HUD CDBG, FEMA, DOE, BEAD jobsPayment-ready BABA evidence package4.3High — statute, FHWA rule, payment-hold specs, advisory incumbents, FCA backdropWINNER
EPCRA Tier II Managed FilingEHS managers, multi-site industrialsAccepted Tier II submission3.6High pricing/outsourcing proof (RMA $1.5–6k) but low intelligence threshold; crowded consultants; adjacent to prior TRI engineReject — commoditized annual form
AES/EEI Export Filing DeskUSPPIs / mid-market exportersITN before departure3.1Verified penalties & filing rules; but forwarders already fulfill; low judgment; weak Sam AltmanReject — incumbent-commoditized
State Auto-Renewal / Click-to-Cancel AuditSubscription SaaS / DTC brandsCancellation-flow compliance pack2.4Federal 2024 rule vacated by 8th Cir (Jul 2025); ANPRM only — softenerDisqualified — unstable federal trigger
Section 45Q Form 8933 Evidence PackCCS project sponsorsAudit-ready sequestration credit file3.4Notice 2026-1 safe harbor real; tiny addressable set; CPA/tax-opinion heavyReject — narrow + licensing drag
Creditor Proof-of-Claim 9037 MonitorBankruptcy creditors / servicersFiled claim + privacy-safe exhibits3.0Adjacent to FileMask (H16) — high dup riskReject — anti-dup

04 Hard disqualifier check

#DisqualifierStatusNotes
1Customer-facing co-pilot / SaaSPassSells completed evidence package + cert chase; specialist is interface.
2Substantial physical laborPassDocument/workflow only; no site crews (site visits optional later, not required).
3Hourly primary pricingPassPer-project package + per-SKU chase fees.
4No path to 50%+ GMPassHigh automation on extraction/chase; specialist minutes concentrated on exceptions.
5Buyer unclearPassGC project controls / compliance manager; owner/grantee as secondary.
6Non-decomposable workflowPassSKU list → classify → chase → validate → package → waive.
7Fully automate regulated judgmentPassSpecialist reviews category/waiver; counsel for FCA/legal opinions; client certifies pay apps.
8Duplicative of prior blueprintPassNo prior BABA/domestic-content slug; distinct from IRA-PWA & certified payroll.
9Illegal / unworkable licensingPassNo PE/attorney license required for documentation assembly; UPL avoided by not giving legal opinions.
10Unverified core demandPassPayment-hold specs + advisory product pages + FHWA rule = verified demand/budget.
11Model progress commoditizes usPassModels help; moat is supplier graph + agency templates + payment-package trust.
12Cannot pilot smallPassOne FHWA/SRF project, 50–200 SKUs, 30-day pilot.

05 Rubric scorecard

Gate 1 — Low trust burden
4.5

Work already outsourced to advisors/consultants; buyers care about unlocked payment and audit file, not the tool.

Gate 2 — Low task-level judgment
4.0

Most steps = extract, classify, chase, completeness check; judgment at category edge-cases and waiver strategy.

Gate 3 — High intelligence threshold
4.0

Agency overlays, product classification, component-cost math, and waiver narratives beat offshore data entry.

Gate 4 — Regulation as moat
5.0

Statutory BABA + 2 CFR 184 + agency rules + payment holds + FCA exposure.

Gate 5 — No physical labor
5.0

Fully remote document/workflow service.

Gate 6 — Sam Altman test
4.0

Better models → faster cert extraction, smarter chase, cheaper packages; supplier graph + agency playbooks compound.

Total gates: 26.5 / 30.

06 Opportunity

IIJA-era
Federal infrastructure FFA still flowing; BABA preference applies to infrastructure FFA generally (not only IIJA) under 2 CFR 184 V
Oct 2026
FHWA 55% manufactured-product component-cost cliff for obligations on/after this date V
Pay hold
CDBG/municipal specs: no progress payment without BABA certs V
$6.8B
DOJ FCA recoveries FY2025 (backdrop for false domestic-origin risk) V
Serviceable wedge (inferred): ~8,000–20,000 mid-market US contractors active on federally assisted civil/water/broadband/energy jobs × 1–4 BABA-heavy projects/year × $8k–$35k package = roughly $0.6B–$2.8B annual serviceable spend if fully penetrated. Treat as order-of-magnitude only; contractor census by funding stream is not a single public series. Inferred

CODE

C — Consumer/buyer trend

FHWA manufactured-products waiver rescission + Oct 2025/2026 phase-in; multi-agency BABA flow-downs (EPA SRF, HUD, FEMA, DOE, BEAD).

O — Opportunity

Project teams lack a production system for product-specific, project-stamped manufacturer certifications and category-correct files.

D — Demand

Payment-withhold language in bid docs; Cherry Bekaert / Free Market America productized offers; contractor trade guidance urging dedicated BABA oversight.

E — Economics

Per-project fees $8k–$35k core; add-on per-SKU chase $75–$250; waiver packets $5k–$20k — outcome-priced against payment risk and rip-and-replace.

07 Evidence quality and source-claim matrix

ClaimLabelSourceConf.Impact
BABA (IIJA §§70901–27) requires iron/steel, manufactured products, and construction materials on covered infrastructure FFA to be produced in the US unless waivedVerifiedOMB M-24-02; 2 CFR Part 184 FR (Aug 23, 2023)HighCore regulatory moat
2 CFR Part 184 is primary OMB guidance; M-24-02 supplements and replaces M-22-11VerifiedWhite House M-24-02 PDFHighRules layer design
FHWA Manufactured Products Final Rule effective Mar 20, 2025; final assembly for obligations ≥ Oct 1, 2025; 55% component cost ≥ Oct 1, 2026VerifiedFHWA Q&A (updated Jan 5, 2026); 90 FR 2932HighWhy-now catalyst
Manufacturer self-certification is the federal documentation model; letters must be product- and project-specificVerifiedUS Made Supply BABA guide; agency sample cert practicesHighDefines deliverable
Some grantees withhold contractor payments until BABA certifications receivedVerifiedWI DEHCR BABA docs summary; Iowa/municipal specsHighBuyer urgency / WTP
Advisory firms sell BABA assessment, cert chase, and documentation servicesVerifiedCherry Bekaert BABA page; Free Market AmericaHighBudget validation
False domestic-origin / Buy America certifications can support FCA liability; FY2025 FCA recoveries >$6.8B; 1,297 qui tamsVerifiedDOJ OPA FY2025 FCA release; Hinckley Allen; Law.comHighRisk narrative (note: $6.8B is all-FCA, not BABA-only)
Mid-market contractors will pay $8k–$35k for DFY BABA packagesInferredAnalog to advisory scopes + payment-hold cost of capitalMedPricing hypothesis — validate in pilots
Serviceable market ~$0.6B–$2.8BInferredContractor × project × fee modelLow–MedSizing only; not a go decision alone
Active LinkedIn/forum contractor complaints volumeUnverifiedNot systematically sampled this runLowDo not rely; use payment-hold + vendor pages instead

08 Why now

  • Verified: FHWA manufactured-products phase-in (Oct 2025 assembly / Oct 2026 55%) expands cert burden beyond melt-and-pour steel culture.
  • Verified: Multi-agency BABA implementation (HUD PIH 2025-06, EPA SRF, FEMA, DOE, BEAD self-cert lists) creates fragmented templates — a production engine wins on agency playbooks.
  • Inferred: Frontier models now reliably extract fields from messy PDFs/emails and draft supplier chase sequences, collapsing COGS vs 2022-era manual desks.
  • Unverified: Whether mid-market GCs will prefer contingency-free fixed packages over expanding internal project-controls headcount — must be tested in pilots.

09 Customer & PMF

ICP$25M–$400M revenue GC or specialty sub with ≥1 active federally assisted civil, water, broadband, or energy project; thin compliance staff.
Economic buyerVP/Director of Project Controls, Compliance Manager, or CFO (payment risk).
UserProject engineer / PE admin / procurement lead who currently chases manufacturer letters.
TriggerNotice to proceed; first pay app rejected for missing BABA docs; FHWA/SRF obligation after Oct 2025/2026 thresholds; owner audit notice.
AlternativesInternal chase; big-4/advisory project; manufacturer marketplace; do-nothing until withhold.
JTBD“Get every covered SKU certified project-specifically so my pay apps clear and my file survives audit — without hiring a BABA clerk.”
WTP evidenceVerified payment holds + verified paid advisory category; exact mid-market ASP inferred.

10 The outcome we sell

Paid outcome: Payment-Ready BABA Evidence Package

  • Deliverable: SKU register with BABA category (iron/steel, manufactured, construction material, excluded/de minimis); project-stamped manufacturer certifications on letterhead; supporting mill tests / assembly location / component-cost attestations as required; completeness score; exception log; optional waiver packet draft for owner submission.
  • Acceptance: ≥98% of covered SKUs have category + cert or documented waiver/exception path; package formatted to owner/agency checklist; specialist sign-off memo.
  • Promise: We assemble the evidence file and chase manufacturers; we do not guarantee agency payment timing or waive legal liability for client false certifications.
  • Exclusions: Legal opinions on FCA exposure; PE stamps; on-site material testing; tax advice; Buy American Act (direct federal procurement) unless scoped as add-on.
  • Rework: Free correction of our classification/packaging errors within 30 days; manufacturer non-response after N chase cycles billed as exception SKUs.
  • Success metric: Pay-app documentation acceptance / owner checklist pass rate; days-to-complete-cert for median SKU.
One-feature MVP wedge
ICP: Mid-market GC on FHWA or EPA SRF project · Trigger: NTP or first pay-app reject · Pain: Missing project-specific manufacturer certs · One-feature MVP: DFY manufacturer-cert chase + payment package for one project · Input: bid schedule / submittal log / POs · Output: cert vault + pay-app exhibit · Human chokepoint: domestic-content specialist category & completeness sign-off · Success: owner accepts package / pay app clears docs · Next ask: multi-project portfolio retainer + waiver desk + supplier pre-qualification graph.

11 Internal AI engine architecture

  1. Intake: Contract BABA clauses, funding agency, bid schedule, submittal register, POs, prior certs, SDS/cut sheets via secure upload / email.
  2. Normalization: SKU identity resolution, manufacturer aliases, unit costs, permanently-incorporated flags.
  3. Retrieval/knowledge: 2 CFR 184 definitions, FHWA 23 CFR 635.410 overlays, agency checklists (EPA/HUD/FEMA/DOE/BEAD), prior gold-standard letters, waiver templates.
  4. AI workbench: Extract product attributes; draft category hypothesis; generate manufacturer chase emails; parse returned letters; flag generic/non-project-specific language; draft waiver narratives.
  5. Deterministic rules: Threshold dates by obligation; iron/steel vs manufactured vs construction material classification tests; de minimis / excluded materials; required letter fields checklist.
  6. Human chokepoint: Specialist approves category edge-cases, accepts/rejects letters, escalates waiver strategy; counsel only if client requests legal opinion.
  7. QA: Dual completeness score; red-team sample of 5% SKUs; hash-versioned vault.
  8. Delivery: Owner-ready PDF/ZIP + pay-app exhibit index; portal link for client (view-only).
  9. Learning loop: Rejected letters → template library; agency comments → checklist updates.
  10. Model portability: Prompt/tooling abstracted; swap LLM providers without changing SOPs.

12 AI-vs-human operations pipeline

AI

Ingest schedule/POs; extract SKUs

Rules

Apply agency date & category tests

AI

Draft chase + parse cert PDFs

Operator

Manufacturer follow-ups / phone

Specialist

Approve categories & letter quality

Client

Signs pay-app certifications

13 Operations as product

Variance killed via: intake checklist (funding agency, obligation date, covered categories); required-evidence list per category; automated completeness gates before specialist review; exception queues (no-response manufacturer, ambiguous permanently-incorporated, multi-category kits); reviewer assignment by agency playbook; confidence scores on letter field extraction; audit trail of every chase; gold-standard letter corpus; red-team sampling; root-cause on pay-app rejects.

14 No-holes quality engine

  • Reject generic “Made in USA” letters lacking project/product identity.
  • Block delivery if obligation-date rule requires 55% component-cost attestation and letter omits it.
  • Separate excluded materials / temporary items from covered permanently-incorporated items with explicit rationale.
  • Never auto-file waivers to agencies — owner/grantee submits; we prepare packet only.
  • Never coach client to certify known-false domestic content (FCA bright line).

15 Pricing, legality, and unit economics

Primary model: Fixed per-project BABA Evidence Package $8,000–$25,000 (≤150 SKUs) / $18,000–$35,000 (151–500 SKUs) + $75–$250 per additional SKU chase + waiver packet $5,000–$20,000. Portfolio retainer optional after 3 projects.

Reject hourly: Misaligns with automation; buyers buy payment unlock, not hours.

Pricing legality: No contingency on federal funds released or penalties avoided (could look like kickback / conflict). Fixed fees only. We do not take a share of project payments. Inferred best practice; confirm with counsel.

COGS element (per 200-SKU project)LaunchDay 90
Model inference + OCR$40–$120$25–$80
Hosting/storage/CRM$30–$60$20–$40
Operator chase labor8–14 hrs ($280–$700)4–8 hrs
Specialist review3–5 hrs ($450–$1,000)1.5–3 hrs
QA sample0.5–1 hr0.4–0.8 hr
Rework buffer8–12% of labor4–7%
Target gross margin45–55%55–70%
Revenue / FTE target$350k–$550k at scale

Automation %: launch ~55% of steps · day 90 ~70% · year 1 ~80% (chase still partly human).

16 Nonlinear scaling plan

  • Throughput: 1 specialist supervises 4–6 active projects/week by day 90 via exception-only review.
  • Cycle time: median SKU cert close 5–12 business days (manufacturer-dependent).
  • Rework <8%; quality failure (owner reject of our packaging) <3%; escalation to counsel <5% of projects.
  • Margin expands as manufacturer graph caches reusable letters (still project-stamped) and agency checklists harden.
  • CAC payback target <3 months on project fees; lead-magnet→consult 8–15%; consult→pilot 25–40%; pilot→paid 60–80% Inferred.

17 Moat & Sam Altman test

Model improvement strengthens OriginVault by cutting extraction/chase cost and improving letter QA. Defensibility = (1) manufacturer response graph + known good signatories, (2) agency-specific package formats owners recognize, (3) specialist trust at payment chokepoint, (4) audit trails that survive Single Audit / OIG sampling.

Commoditization threat: Procore/Autodesk or LCPtracker-class construction platforms add “BABA module” self-serve. Counter: win on DFY chase + payment-package SLA + specialist sign-off — the labor owners already refuse to staff.

18 Buyer-specific go-to-market

Primary motion: founder-led outbound to project controls + channel via municipal engineers, SRF program consultants, and construction CPAs — not creator waitlist-first. Content/AEO supports inbound but does not replace outbound.

First 50 prospectsGCs with active FHWA/SRF/CDBG bids in 3 target states; specialty mechanical/electrical subs on those jobs.
TriggersNTP; pay-app reject; obligation after Oct 2025/2026; owner BABA kickoff meeting.
Outreach wedge“We’ll return a payment-ready BABA cert vault for Project X in 15 business days — fixed fee.”
Credibility assetAnonymized redacted package + agency checklist crosswalk.
Sales cycle2–6 weeks for single project; longer for portfolio.

Distribution proof table

ChannelWhy ICP reachableFirst angleConv. assumptionProofMeasureFollow-up
LinkedIn outboundProject controls titles publicPay-app withhold teardown3–6% replyTrade articles on BABA burdenReply→meetingDiagnostic offer
SRF/municipal engineer referralsThey see incomplete filesReduce their review cycles1 in 5 warm introsSpec payment-hold languageIntro→pilotCo-branded checklist
AGC / AWWA chaptersContractor densityOct 2026 55% workshop10–20 leads/eventFHWA ruleEvent→waitlistOffice hours
Search / AEO“BABA certification letter requirements”Field checklist lead magnet2–5% to consultQuery demand inferredMagnet→booked callSequence
Construction CPA / advisory partnersAlready in trust loopWhite-label packagePartner-dependentCherry Bekaert categoryReferral countRev share

First 30 days of content

  • 10 posts: generic vs project-specific letters; iron/steel vs manufactured; Oct 2026 55% cliff; pay-app withhold examples; de minimis myths; kit classification; permanently-incorporated gray zones; waiver timeline reality; FCA risk without legal advice; supplier chase scripts.
  • 3 teardowns: redacted bad letter; incomplete pay-app exhibit; misclassified HVAC unit.
  • 2 magnets: “BABA Letter Field Checklist” + “Obligation-Date Rule Card (FHWA).”
  • 1 live review: office hours on 10 anonymous SKUs.
  • 1 outbound diagnosis template: 12-line project gap memo.

Lead magnet & waitlist

Free BABA Payment-Risk Diagnostic: upload bid schedule sample → AI draft category map + missing-cert estimate → specialist 20-min review. Waitlist CTA for pilot cohort (cap 8). Qualifies when ≥30 covered SKUs and active federal funding.

19 Pilot design and early-demand trap

  • Pilot cap: 8 projects / 5 customers max.
  • Success: ≥90% SKU completeness without heroics; median chase ≤10 business days; owner accepts package; COGS tracked per SKU.
  • Refuse: custom on-site inspections, unrelated Davis-Bacon payroll, full ERP builds.
  • After 5: harden intake + letter field QA. After 10: harden SOPs/templates. After 20: pause until COGS/rework measured.

20 Competitive landscape

PlayerTypeGap vs OriginVault
Cherry Bekaert / big-4 advisoryServicesExpensive, slow for mid-market single projects
Free Market AmericaSourcing/docs supportCloser; differentiate on AI production system + payment-package SLA
US Made Supply / manufacturer marketplacesProduct discoveryHelps find products; does not run DFY chase/file for contractor
Internal project controlsDo-it-yourselfLabor shortage; inconsistent letter quality
Construction SaaS modulesSoftwareCustomer-operated; no manufacturer chase engine

Anti-duplication

Not IRA-PWA (labor/payroll). Not certified payroll. Not generic “AI for construction.” Narrow wedge = domestic-content evidence production for payment and audit under BABA/agency overlays.

21 Regulation, compliance, and licensing boundary

  • AI may: extract, classify hypotheses, draft chase, parse letters, assemble packages, draft waiver narratives.
  • Operators may: contact manufacturers, track exceptions, format exhibits.
  • Specialists must: approve categories/letter sufficiency; decide waiver-prep vs escalate.
  • Licensed counsel (client’s or ours, fractional): FCA exposure opinions, disputed legal interpretations — not routine packages.
  • Client must: sign contractual pay-app certifications and remain legally responsible for truthfulness.
  • We must not claim: we “certify” products as BABA-compliant to the government; we prepare evidence and specialist QA.
  • Privacy: project docs may include pricing — SOC2-minded controls; no unnecessary PII.

22 Compact founding team and expert map

RoleWhyFT/FractionalTiming
Domain specialist (ex-GC compliance / SRF)Chokepoint trustFT or heavy fractionalDay 0
Ops leadSOPs, chase SLAsFTDay 0
Automation/AI engineerEngineFT or fractionalDay 0
Construction counselBoundary opinionsFractionalAs needed
Sales/channelOutbound + partnersFractional→FTAfter 3 pilots
QA ownerRed-teamShared with opsDay 0

What the human expert actually does

TaskLicenseMin launchMin d90Automation pathQuality riskCannot automateAudit trail
Category edge-case approvalNone (competency)2512Rules+model suggestMisclass → pay rejectAmbiguous kits/fixturesDecision log
Letter sufficiency sign-offNone2010Field extraction QAGeneric letter acceptedNovel manufacturer formatsVersioned PDF hash
Waiver strategyNone / counsel if legal4025Draft narrativeWrong waiver typeOwner politicsPacket versions
Client briefingNone3020Auto summaryOverpromiseTrustCall notes

23 Exhaustive risk register

Agency template fragmentation
Likelihood: MedImpact: HighEvidence: VerifiedOwner: Ops

Mitigation: Maintain per-agency playbooks; start FHWA+SRF only

Leading indicator: Checklist drift tickets

Manufacturer non-response
Likelihood: HighImpact: HighEvidence: VerifiedOwner: Ops

Mitigation: Multi-channel chase + client escalation clause

Leading indicator: % SKUs >14d open

FCA association / client false cert
Likelihood: MedImpact: CriticalEvidence: VerifiedOwner: Compliance

Mitigation: Refuse known-false; engagement letter; counsel escalation

Leading indicator: Escalation count

Misclassification of products
Likelihood: MedImpact: HighEvidence: InferredOwner: Specialist

Mitigation: Specialist dual-review on low confidence

Leading indicator: Owner rejects

Construction SaaS launches module
Likelihood: MedImpact: MedEvidence: InferredOwner: CEO

Mitigation: Double down on DFY chase SLA

Leading indicator: Competitor launches

Political softening of BABA
Likelihood: LowImpact: HighEvidence: UnverifiedOwner: CEO

Mitigation: Diversify agencies; monitor MIAO

Leading indicator: Waiver policy changes

Scope creep into Davis-Bacon/PWA
Likelihood: HighImpact: MedEvidence: InferredOwner: Sales

Mitigation: Hard SOW exclusions

Leading indicator: Custom hours

Pricing below COGS on hard projects
Likelihood: MedImpact: MedEvidence: InferredOwner: Finance

Mitigation: SKU-band pricing; kill switch

Leading indicator: GM per project

Data security of bid pricing
Likelihood: MedImpact: HighEvidence: InferredOwner: Eng

Mitigation: Encryption, access logs, DPA

Leading indicator: Incidents

Talent: scarce specialists
Likelihood: MedImpact: HighEvidence: InferredOwner: Ops

Mitigation: Train operators; gold examples

Leading indicator: Hire funnel

Waiver timelines blow SLAs
Likelihood: HighImpact: MedEvidence: VerifiedOwner: Ops

Mitigation: Set expectations; separate waiver SKU

Leading indicator: Waiver cycle days

Over-reliance on inferred TAM
Likelihood: MedImpact: MedEvidence: InferredOwner: CEO

Mitigation: Pilot revenue before scale spend

Leading indicator: Paid pilots

24 Tech stack & build plan

  • Week 1–2: Airtable/Notion vault + Gmail/Outlook chase sequences + Claude/GPT extraction scripts + checklist PDF templates.
  • Week 3–6: Ingest pipeline, manufacturer CRM, letter parser, completeness scorer, client view-only portal.
  • Week 7–12: Agency playbook pack (FHWA, EPA SRF), waiver drafter, analytics on chase conversion.
  • Avoid building a full construction ERP; integrate via CSV/API later.

25 Metrics & KPIs

Throughput projects/FTE · cycle time per SKU · rework % · gross margin · COGS/project · revenue/FTE · escalation % · automation % · evidence completeness % · quality failure % · customer acceptance % · pilot conversion % · manufacturer response rate · pay-app doc pass rate.

26 What could kill this

  • Federal rollback that restores broad manufactured-product waivers and collapses urgency.
  • Incumbent platform bundles “good enough” DFY via partner network at scale.
  • Inability to get manufacturers to respond — if chase conversion <40%, model fails.
  • One FCA scandal involving a client who ignored our refuse-to-falsify policy (reputational contagion).

27 90-day validation and launch plan

WindowActionsKill/continue criteria
Days 1–7SOW + engagement letter; FHWA/SRF checklists; 20 prospect list; diagnostic landing pageCannot name 20 real projects → pause
Days 8–303 design partners; manual fulfillment; instrument COGS; publish 10 posts<1 paid pilot LOI → revisit ICP
Days 31–60Complete 3 packages; measure chase conversion; harden letter QAChase <40% or GM <35% → redesign
Days 61–90Cap at 8 pilots; pricing test; partner conversations; decide scaleOwner acceptance <80% → no-go scale

7 / 30 / 90 day snapshots

7-day: offer page, checklist magnet, 50 outbound, counsel-reviewed engagement letter.
30-day: first paid package in flight, weekly feedback loop, COGS dashboard.
90-day: 5–8 completed packages, playbooks for 2 agencies, go/no-go on hiring second specialist.

Final recommendation: Proceed to blueprint execution. OriginVault clears evidence threshold on regulation, payment pain, and incumbent budget proof, with a narrow DFY MVP and explicit licensing boundary. Validate chase conversion and mid-market ASP in the first 8 pilots before scaling.

28 Sources

  1. OMB M-24-02 Buy America Implementation Guidance
  2. Federal Register — OMB 2 CFR Part 184 final guidance (Aug 23, 2023)
  3. FHWA Manufactured Products Final Rule Q&A (updated Jan 5, 2026)
  4. FHWA Buy America Requirements for Manufactured Products final rule (90 FR 2932)
  5. FHWA newsroom — manufactured products waiver updates
  6. Cherry Bekaert — BABA compliance rules & services
  7. Free Market America — BAA/BABA documentation support
  8. US Made Supply — Buy America compliance guide for contractors
  9. Wisconsin DEHCR — BABA compliance documents / payment holds
  10. HUD PIH 2025-06 — BABA implementation notice
  11. Hinckley Allen — BABA exposure / FCA tips for contractors
  12. DOJ — FCA settlements & judgments exceed $6.8B FY2025
  13. Construction Business Owner — what contractors should know about BABA
  14. DOE — Build America, Buy America program page
  15. Law.com — growing fraud risk of domestic-origin claims

Generated 2026-07-09T18:02:00Z. Claims labeled Verified / Inferred / Unverified in-matrix. This blueprint is not legal advice.