RetinaScan

• Live frontier: multimodal diagnostic models, FDA SaMD pathways, RWE evidence loops.
• Tech feasibility score 69/100 — driven by 23% sector innovation rate.
• Capital intensity 60% sets the R&D burn profile.

• Scientific claims unverified without a technical deep-dive / reference customers.
• reimbursement dependency and long clinical validation cycles.

• Market factor 66/100; unit-economics factor 54/100.
• Reference gross margin ~60%; validate against actuals.
• Execution signal 64/100 — Quantified traction: $480k MRR..

• TAM/SAM/SOM and CAC/LTV unconfirmed — require a live data room.
• Sector benchmarks are directional, not company-specific.

• The 66/100 is generously anchored to a ~$420B digital-health TAM at 23% CAGR. Push back: RetinaScan does not sell into $420B — the relevant serviceable market is US diabetic-retinopathy screening. ~38M US diabetics, ~60% under-screened, reimbursed CPT 92227-92229 at roughly $30-65/screen implies a bottom-up SAM closer to $1-3B, not hundreds of billions. Score is directionally right on tailwinds, w
• Demand is reimbursement-elastic, not price-elastic: PCP adoption follows whether the screen is billable and margin-positive in-visit. IDx-DR/LumineticsCore already set the CMS precedent (autonomous AI DR screen reimbursed), which de-risks the payer path but also proves an incumbent is 5+ years ahead with FDA De Novo clearance — a disclosed, well-capitalized competitor the 58/100 competitive headro
• Economic rents here accrue to the regulatory license plus the distribution/EHR-integration layer, not the smartphone lens (commoditizable hardware). Moat at 63 is fair-to-slightly-generous at seed: breakthrough-device is only *filed*, not granted, and 91% sensitivity in 900 patients is a strong pilot but below the ~87-90% sensitivity AND ~90% specificity bar CMS/FDA-cleared comparators cleared — s
• $40k MRR across 18 clinics = ~$2.2k/clinic/month, a plausible SaaS+per-screen mix but a tiny base; unit economics (LTV/CAC, 60% capital intensity) untested. The composite's 61.7 'watch' is the honest read: real evidence, unproven that it compounds past the incumbent.

• Reimbursement dependency (structural): CMS/private payer coverage for a second autonomous DR tool is not guaranteed; without a durable CPT/NTAP economics, PCP demand collapses since clinics won't screen at a loss.
• Incumbent moat asymmetry: a cleared, reimbursed autonomous competitor can bundle or undercut; RetinaScan must beat specificity and total cost-of-workflow, and specificity was not reported — a material omission for a diagnostic.
• Long validation + FDA cycle burns the $5.5M before revenue inflects; at ~$480k ARR run-rate and 60% capital intensity, runway to clearance and payer contracts likely requires a bridge, raising down-round risk.

• The 42 score reads regulatory intensity (90%) as pure drag; I'd push back on framing. For a SaMD diagnostic, FDA clearance IS the durable moat the Moat factor credits (ceiling 82). The 7% weight on 'headroom' understates that clearance, once obtained, converts drag into a 3-5yr competitor barrier. Net: score is defensible as a burn/timeline read, too pessimistic as a value read.
• Device pathway is unconfirmed. 'Breakthrough-device designation filed' — not granted; De Novo (novel diagnostic, likely) runs ~12-18mo and $1-3M vs a predicate-based 510(k). At $40k MRR the 18 clinics are almost certainly running under a research/adjunct-to-specialist framing, not autonomous marketed diagnosis. Confirm current commercialization is not off-label promotion — that is an FTC §5 / FDA
• AI/algorithm risk is real and under-scored: an ML diagnostic must lock its algorithm or file a Predetermined Change Control Plan; 91% sensitivity in a 900-patient study is single-site and needs multi-site RWE for generalizability (bias across skin tone/camera hardware). HIPAA BAAs with all 18 clinics and CCPA/CPRA on any CA patient data are table stakes — verify executed, not assumed.
• Structure the round to price regulatory milestones: Reg D 506(b) priced round or post-money SAFE, 1x non-participating pref, board observer + information rights, and a tranche/milestone release tied to De Novo grant or breakthrough designation. Add reps on FDA correspondence, absence of warning letters, and clean IP assignment from clinical/academic founders (retina algorithms often trace to unive

• Clearance failure or delay: if De Novo stalls past 18mo, the entire moat thesis and reimbursement path collapse; $5.5M seed likely funds to grant only if burn is disciplined — no runway buffer disclosed.
• Reimbursement dependency (flagged as structural risk): even with clearance, no CPT/coverage pathway means clinics pay out-of-pocket — the $40k MRR does not prove payer economics and LTV/CAC was not disclosed, so the revenue moat is unstressed.
• Off-label / promotional-claims exposure now: pre-clearance marketing of a diagnostic claim invites FDA/FTC action that would poison the pending submission and any acquisition due diligence.