What changed
FACT: A federal Medicaid rule, 'Community Engagement Requirement for Certain Individuals' (Federal Register 2026-11094, published 2026-06-03), creates a new regime requiring certain Medicaid beneficiaries to document qualifying activities (work/volunteer/education hours) and every state to stand up verification/reporting systems. INFERENCE: this converts millions of enrollees into a forced-filer class who must submit monthly or lose coverage.
Why now
FACT: cheap schema-defined structured extraction from arbitrary documents is newly available as a single API call (Context.dev, YC S26). HYPOTHESIS: combined with the just-finalized rule and staggered 12+ month state rollouts, there is a build-ahead window before incumbents (state MCO vendors, eligibility-system integrators) wire this in.
Converging signals
Three signals meet: (1) the federal community-engagement rule forcing a defined filer class; (2) each state's obligation to build a verification/reporting portal on its own forms/cadence; (3) near-zero-cost document extraction that makes turning messy stubs/logs into a submission economically viable per enrollee.
Customer pain
FACT (from rule shape) + INFERENCE: coverage loss under work requirements is driven largely by paperwork confusion and missed deadlines, not actual ineligibility — a well-documented pattern from prior state work-requirement experiments (Arkansas). The enrollee's pain is silent disenrollment; the BUYER's pain is churn.
Who pays
Managed-care organizations (MCOs) paid capitation per retained member, FQHCs/community health centers, and navigator/enrollment-assister nonprofits funded to prevent improper coverage loss. Model: per-enrollee-retained, per-seat for navigators, or white-label to the MCO. The beneficiary is NOT the buyer.
Solved today
HYPOTHESIS: today it's manual — caseworkers, navigators, VAs, and clinic staff chasing members for documents, plus generic state portals the enrollee must self-navigate. No focused autofiler exists yet because the rule is weeks old.
Why current solutions are bad
Manual outreach doesn't scale to monthly cadence across a member panel; the state portal assumes the enrollee correctly interprets qualifying-activity rules and formats hours — precisely where confusion causes disenrollment.
Proposed product
A white-label mobile + API product: enrollee photographs stubs/logs → extraction pulls hours/wages/dates → a per-state rules layer maps them to that state's qualifying-activity definitions and reporting cadence → generates the state-format submission + deadline reminders. Sold to MCO/FQHC/navigator as a retention tool, dashboard for their staff.
MVP version
Pick ONE state that has published finalized qualifying-activity rules. Encode that one state's rules. Wire Context.dev extraction for stubs + volunteer logs. Output the completed state-format report (PDF/portal-ready) for staff review. Run the KILL TEST first: real redacted stubs → does a caseworker accept the output uncorrected?
30-day build
Run the kill test on one state with a batch of real redacted documents; measure extraction accuracy and caseworker-acceptance rate. Interview 5-8 MCO retention leads / FQHC enrollment managers / navigator programs to confirm willingness to pay per retained member. Encode state #1 rules.
60-day build
Build the staff-review dashboard + reminder pipeline; sign one pilot channel partner (MCO or FQHC) as a paid or LOI'd design partner; validate the submission is portal-acceptable for that state.
90-day revenue plan
Convert the pilot to a paid per-enrollee or per-seat contract; add a second state as its rules finalize. First revenue via the channel partner, not direct-to-enrollee.
Distribution path
Direct outreach to MCO member-retention/quality teams, FQHC enrollment managers, and state navigator/assister networks — a small, reachable buyer set, not open-market consumer acquisition. Association channels (state primary care associations) for FQHCs.
Pricing hypothesis
Per-enrollee-retained (e.g. $3-8/member/month for the covered panel) or per-navigator-seat ($100-300/seat/mo), white-label option at a platform fee. Anchor against the cost of a churned member to the MCO.
Technical difficulty
Moderate. Extraction is a solved API call; the real work is the per-state rules layer and state portal/format integration (fragile, changes per state), plus handling PII securely.
Legal / regulatory risk
Handling enrollee PII/PHI-adjacent data (HIPAA exposure if partnered with covered entities) requires BAAs and secure storage. No licensure required to build the tool. Not medical advice.
Platform dependency
Low platform-policy risk — submits to government systems, no app-store gatekeeper who can deplatform. Dependency risk is on state portal stability and on the extraction API vendor (mitigate with a fallback extractor).
Founder fit
HIGH. This is the founder's proven shape: a federal mandate compels a class to file to government systems, and he has shipped exactly this (FMCSA ELDT portal autofiler charging per upload). Public-records + automation + AI-workflow strengths all apply.
Breakout potential
HIGH if the wedge holds: 50 near-identical state markets to replicate into once state #1 works, monthly recurring cadence, and a captive forced-filer population numbering in the millions.
Final recommendation
PURSUE, gated on the kill test. This is a top-tier fit for the founder's primary forced-filer thesis with a fresh federal mandate, a captive monthly-filing population, a reachable non-enterprise buyer set (FQHCs/navigators/regional MCOs), and a 50-state replication path — but the whole thesis hinges on extraction+rules producing caseworker-acceptable submissions and on the rule surviving litigation. Validate both before heavy build.
Next action
Obtain one state's finalized qualifying-activity rules + a batch of real redacted pay stubs/volunteer logs and run the extraction→state-format submission kill test; in parallel, book 5 discovery calls with MCO retention / FQHC enrollment / navigator leads to confirm per-retained-member willingness to pay.