What changed
FACT (cited): In June 2026 FDA issued a cluster of final classification orders creating new Class II device categories with special controls — radiological ML-based quantitative imaging software with a Predetermined Change Control Plan (6/17), opioid-induced-oxygenation-impairment monitors (6/30), a preeclampsia prognostic test (6/26), an ingestible GI blood-detection capsule (6/17), Bacillus IVD (6/22), POC SARS-CoV-2 device (6/11), and a breast-implant suction retrieval system (6/26). Each order codifies special controls that every marketer of that device type must conform to.
Why now
The obligated class is forming right now: these product codes and their special-controls checklists did not exist before June 2026, so no incumbent has a productized, code-specific offering yet. The PCCP mechanism in the ML-imaging classification creates a genuinely novel recurring documentation duty (evidence per model revision), not just a one-time premarket burden. HYPOTHESIS: firms entering these codes are mostly startups too small for enterprise QMS suites.
Converging signals
Seven FDA Class II classification orders in a 3-week window (all cited in demand_evidence), several targeting software/ML and diagnostics — the segment where small firms cluster. FDA's public Establishment Registration & Device Listing database enumerates every marketer by product code, giving a self-updating lead list. No PAIN or HIRING/SPEND evidence was retrieved — demand rests entirely on the FORCED BUYER mandates.
Customer pain
HYPOTHESIS (not evidenced in input): a 5–50 person device/SaMD startup must continuously demonstrate conformance to its code's special controls — software documentation, bench/clinical performance records, labeling conformance, MDR files, and for ML imaging a documented PCCP change log per model update — with the sanction being misbranding/adulteration (market removal or import refusal), not a fine. Today that lives in scattered Google Drive folders assembled in panic before an audit or 510(k).
Who pays
Regulatory affairs / quality lead (often a founder wearing that hat) at small firms listed under the new product codes and adjacent De Novo-derived codes. The FDA registration database provides names and addresses. CAVEAT (FACT about how De Novo works): a brand-new code often has only the original De Novo grantee marketing under it at first; the 20+ firm count in the testable prediction is UNVERIFIED and is the single biggest open question.
Solved today
eQMS suites (Greenlight Guru, Qualio) at $10k–$30k+/yr, regulatory consultants at $200–$400/hr, or spreadsheets and shared drives. HYPOTHESIS: none offers a checklist keyed to a specific June-2026 product code's codified special controls, and none has a PCCP update-log workflow productized yet (Ketryx is closest).
Why current solutions are bad
eQMS platforms are generic QMS scaffolding priced above seed-stage budgets; consultants are one-time engagements that decay immediately; neither maintains a living, code-specific evidence file mapped line-by-line to the codified special controls, and the PCCP per-model-revision log is too new for templated support.
Proposed product
Evidence Locker: per-product-code subscription. Onboarding ingests the code's codified special controls into a structured checklist; the app maintains the evidence file (document slots, review dates, versioned PCCP change log entries per model release, MDR file tracker, annual-registration deadline watcher), produces an auditor/510(k)-ready dossier export, and monitors the Federal Register for amendments to the code's controls.
MVP version
One product code only — the radiological ML imaging + PCCP code (largest plausible startup population, most novel recurring duty). Static special-controls checklist parsed from the order, document upload slots with status, PCCP change-log form generating a timestamped PDF record per model version, deadline reminders. Buildable solo in 45–60 days with AI-assisted development; no PHI, no device data touches the system.
30-day build
Validation before build: pull FDA Establishment Registration & Device Listing files, count firms under the new codes plus adjacent De Novo ML-imaging codes (QIH, MYN family). Read both orders' special-controls sections to confirm recurring vs premarket-only duties. Cold-email 30 listed firms offering a free 'special-controls gap checklist' for their code. Kill criteria: <15 reachable firms or <3 demo calls booked.
60-day build
If validated: build the single-code MVP, convert the gap-checklist leads into 3–5 design partners at $99/mo founding price. Publish the gap checklist as SEO/LinkedIn content — 'What the June 2026 ML imaging classification actually requires' — to catch firms Googling the new code.
90-day revenue plan
Target 8–12 paying subscriptions at $199–$399/mo/product code (~$2–4k MRR) plus $1,500 one-time 'dossier setup' engagements. Add the opioid-oxygenation-monitor code as the second template to prove the repeatable per-code playbook.
Distribution path
The FDA registration database IS the distribution list — every obligated firm, with address, updated continuously. Direct outreach demonstrating a pre-built gap analysis for their exact code (demonstrated-value sales, matching founder's style), plus RAPS forums and LinkedIn regulatory-affairs groups. New classification orders become a recurring content/lead trigger.
Pricing hypothesis
$199–$399/mo per product code per firm, $1,500 one-time dossier setup, consultant white-label tier later. Anchored far below eQMS ($10k+/yr) and below one consultant-day.
Technical difficulty
Low-moderate: document management, checklists, versioned logs, PDF export, Federal Register API watcher. No integrations with hospital or device systems. The hard part is regulatory-content accuracy, not code.
Legal / regulatory risk
Moderate and real: the product is adjacent to regulatory advice. Must position as document-management tooling with the codified text quoted verbatim, not as advice; an error that lulls a customer into non-conformance creates liability and reputation risk in a domain where the founder has no credential. E&O insurance and a consultant reviewer advisable — affordable given capital.
Platform dependency
Low. Public FDA data and Federal Register feeds; no marketplace gatekeeper.
Founder fit
Mixed — weaker than it looks. The shape rhymes with the FMCSA ELDT win (read a mandate, find the obligated class, monetize compliance) and the public-roster lead list plays to his public-records strength. BUT this is not a portal-submission play: there is no per-filing government upload to automate (annual FURLS registration is trivial); the product is ongoing documentation SaaS, which lacks the transactional wedge. And the founder profile explicitly avoids heavily regulated medical products — selling compliance tooling INTO med-device is softer exposure than making a device, but buyers will ask 'who are you in FDA regulatory?', and he has no answer. Fit is moderate (5/10), well below his government-portal pattern.
Breakout potential
Good if validated: FDA issues classification orders continuously, so the per-code template is a repeatable factory; PCCPs are spreading beyond radiology across AI-enabled devices, and the PCCP log could become the wedge into broader AI-device lifecycle compliance. Expansion is real but each step deepens exposure to the med-device trust problem.
Final recommendation
CONDITIONAL — do not build yet. The mandate evidence is real and the pattern is elegant, but the two load-bearing unknowns (obligated-class size; willingness of med-device buyers to trust a non-credentialed solo vendor) are both cheap to test. Spend 2 weeks and near-zero dollars on the listing-count pull and 30 cold contacts. If 15+ firms exist and 3+ take a call, fund the 60-day single-code MVP; otherwise archive the template and re-run it against the next classification cluster in a less trust-gated domain.
Next action
Download FDA's Establishment Registration & Device Listing public files and count distinct firms listed under the new June-2026 product codes plus adjacent De Novo ML-imaging codes; simultaneously read the codified special-controls text of the 6/17 ML-imaging and 6/30 opioid-monitor orders to confirm the duties are recurring rather than premarket-only.