What changed
FACT: USAspending shows massive still-open FEMA PA obligations flowing through state pass-throughs — $1.53B COVID PA to Florida DEM (DR-4834 line), $2.89B (DR-4673 FL), $4.94B (DR-4486 FL), $2.35B (NC), $1.60B (GA), $1.80B (PR), $2.33B (PA/PEMA) — all reimbursement-based, meaning every downstream subrecipient must produce documentation to get and keep the money. INFERENCE: pandemic-era PA (declared 2020) is now deep in the closeout/audit window, the phase where documentation gaps become de-obligations and clawbacks.
Why now
PA is reimbursement-after-the-fact: the state (FDEM etc.) passes money to thousands of subrecipients who must submit Project Worksheet cost documentation, quarterly reports, and closeout packages via FEMA Grants Portal and state systems like FloridaPA.org. COVID projects are hitting closeout simultaneously across every state, and FEMA/OIG closeout audits routinely claw back funds for missing payroll records, procurement documentation, or duplicate-benefit issues. The pain peaks NOW, in a compressed multi-year window. HYPOTHESIS: most small subrecipients (volunteer fire departments, small towns, PNPs) have no grant-management software and are tracking this in folders and spreadsheets.
Converging signals
(1) FACT: $1.53B FEMA PA COVID award to Florida DEM naming reimbursement to SLTT governments and PNPs (usaspending.gov/award/ASST_NON_4834DRFLP00000001_070); (2) FACT: six-plus additional multi-billion PA awards across FL/NC/GA/PR/PA in the same evidence set, proving this is a national repeating pattern, not one grant; (3) FACT: a defined portal pair (FEMA Grants Portal + FloridaPA.org) and a defined paperwork burden (PWs, cost docs, quarterlies, closeout). Rule + filer class + portal converge — the founder's exact thesis shape.
Customer pain
A reimbursement-based grant means the subrecipient has ALREADY spent the money; failing closeout means writing a check back to FEMA. Small subrecipients must reconstruct 2020-2022 payroll, force-account labor/equipment logs, invoices, contracts, and procurement compliance (2 CFR 200.317-327) years after the fact, formatted to FEMA's documentation expectations. INFERENCE: the person doing this is a town clerk or finance director with no PA training, or a consultant billing hourly/percentage.
Who pays
Primary: small-to-mid subrecipients — municipalities, counties, school districts, special districts (fire/water), hospitals, and private non-profits in FL first (largest open pipeline in evidence), paying per closeout package or per Project Worksheet. Secondary: small PA consultants and CPA firms serving these entities, who buy it as a workbench to serve more clients. Note: PA Category Z management costs make some of this spend FEMA-reimbursable for the subrecipient — the tool can literally be paid for with FEMA money (HYPOTHESIS to verify per program).
Solved today
Large subrecipients hire disaster-recovery consultancies (Hagerty, Tidal Basin, IEM, Witt O'Brien's) that bill hourly or as a percentage of the award. Small ones use spreadsheets, shared drives, and the state's PA field liaisons. Generic grant-management platforms (AmpliFund, eCivis, Neighborly) target the GRANTOR/state side or full grant lifecycle, priced and sold as gov-enterprise software.
Why current solutions are bad
Consultants are proof of spend but are unaffordable below a project-size threshold — a $150k PW cannot absorb a $30k consulting engagement. Enterprise grant suites require procurement and implementation the small subrecipient will never do. Nothing in evidence is purpose-built for the narrow job: 'assemble THIS closeout package to FEMA's documentation checklist and flag what will get clawed back.' That gap between spreadsheet and consultant is the wedge.
Proposed product
PA Closeout Vault: upload/dump documents (payroll exports, invoices, contracts, timesheets), the tool classifies them against FEMA's PA documentation requirements per category (A-G, Z), maps them to Project Worksheets, produces a gap report ('PW 1234 is missing procurement history and equipment logs — clawback risk'), and assembles an indexed, audit-ready closeout binder formatted for Grants Portal upload. AI does classification and gap-detection; deterministic checklists (from the PA Program and Policy Guide) do compliance mapping. Optionally a done-for-you assembly service at first — service revenue funds the product.
MVP version
A single-declaration, Florida-only version: intake form + document upload, checklist engine encoding FEMA's closeout documentation requirements for Categories B (emergency protective measures — the COVID category) and Z, gap report PDF, assembled closeout package as indexed PDF/ZIP matching Grants Portal structure. No portal API integration needed at MVP — output is what a human uploads. Founder has already built portal-submission automation (FMCSA ELDT) and can add FloridaPA.org/Grants Portal submission assistance later.
30-day build
Encode the PA closeout documentation checklist from the PA Program and Policy Guide; build intake + classifier + gap-report on the existing AI stack; pull FDEM's public list of applicants/PWs for the COVID declarations (state PA sites publish obligated-project data) to build a target list of small subrecipients with open projects; interview 5-10 town clerks/finance directors and 3-5 independent PA consultants.
60-day build
Run 3 pilot closeouts (paid or heavily discounted) for FL PNPs or small municipalities sourced from the target list; iterate the gap report until a former state PA reviewer (contract advisor, ~$2-5k) signs off that it matches what closeout reviewers reject on; publish 2-3 'what gets clawed back at COVID PA closeout' briefs as the demand-gen engine.
90-day revenue plan
Charge $1,500-$5,000 per closeout package (or $250-$500 per PW) — under micro-purchase thresholds, so no procurement process; sell the consultant workbench at $500-$1,000/mo to small PA consultants. 10 closeout packages ≈ $25-40k. Revenue in 90-150 days is plausible because the buyer's alternative is a clawback measured in six figures.
Distribution path
Direct outreach to the public list of subrecipients with open PWs (state PA portals and FEMA's OpenFEMA PA datasets enumerate applicants and project counts — a ready-made lead list, founder's public-records strength); Florida League of Cities / association newsletters; the clawback-focused content briefs; partnerships with small CPA firms doing single audits (2 CFR 200 Subpart F) who see the documentation mess first.
Pricing hypothesis
Per-closeout package $1,500-$5,000 scaled by PW count; per-PW $250-$500; consultant seat $500-$1,000/mo. Deliberately under micro-purchase threshold to avoid procurement. Possible framing: 'reimbursable as PA management costs' (verify before claiming).
Technical difficulty
Low-to-moderate: document classification + checklist engine + PDF assembly is squarely within AI-assisted solo build capability. The hard part is domain fidelity — encoding what FEMA closeout reviewers actually reject — which is bought via a contract SME, not engineered.
Legal / regulatory risk
Low-moderate: the tool organizes and formats; it should not certify compliance (avoid practicing law/accountancy claims). Disclaimers + 'preparer, not certifier' positioning. No license required to operate. Handling municipal payroll data requires sane security but no special regime.
Platform dependency
None that can deplatform: FEMA Grants Portal and FloridaPA.org are government systems (per system guidance, no platform_policy_risk applies). Risk is program-lifecycle, not platform: COVID PA closeouts are a finite wave — mitigated because PA recurs with every declared disaster (hurricanes guarantee FL pipeline) and the same engine extends to HMGP/CDBG-DR closeouts.
Founder fit
Near-maximal. This is the proven ELDT shape: government program compels a defined class to submit documentation into a portal; founder builds the paperwork/submission layer and charges per transaction. Adds his public-records skill (mining OpenFEMA/state PA data for the lead list) and fire-service background (native credibility with emergency-management buyers and Category B language). Lesson 'government-portal mandate opportunities fit this founder best' (conf 0.79) applies directly.
Breakout potential
Strong replication path: 50 states run the same federal program on near-identical rules (FL → NC → GA → PR from this very evidence set), then category expansion to HMGP, CDBG-DR, and future declarations. Every hurricane season refills the pipeline. Ceiling is a durable niche compliance firm ($1-5M/yr), not venture-scale — which matches the founder's constraints.
Final recommendation
PURSUE with a validation gate. The forced-buyer structure is genuinely strong ($1.5B+ per state, reimbursement-conditional, defined paperwork, defined portals) and it is the founder's proven ELDT shape almost exactly. But all evidence here proves money and obligation, not willingness of SMALL subrecipients to pay a third party — so the gate is 10 discovery calls with FL subrecipients holding open COVID PWs within 30 days. If ≥3 say they fear closeout/clawback and have no consultant, build. If they all point to free FDEM hand-holding, kill and redirect the engine at the same shape with a harder deadline.
Next action
Pull the FL COVID PA applicant/project list (OpenFEMA PA-funded-projects dataset + FloridaPA.org public data), filter to subrecipients with <$5M obligated and multiple open PWs, and book 10 discovery calls with their clerks/finance directors this month.