Every discovered stock, sorted by velocity score V.
V = risk-adjusted return per day to our exit · Q is a gate (must pass, not a rank input).
The score is computed by the tool over gathered pieces — AI roles never do math.
Exit window = 2–4 weeks before the catalyst binary. Never hold through the binary.
POST /ask with 7 operational business roles. Each role gathers independent pieces. The Coordinator outputs a structured JSON block at the end — those field values are written back to the candidate store automatically. The tool then re-scores.
Scout — Candidate Discovery
The Scout role searches multiple public sources (FDA calendar, SEC EDGAR, BioPharmCatalyst, ClinicalTrials.gov, market data) to discover 3–7 new small/mid-cap biotech candidates. It fills all scoring fields and hands structured JSON back; the UI writes each record directly to the store.
Enrichment Leaderboard
Sorted by V (velocity) descending. Q is a quality gate — names below q_min or with unknown gate inputs are parked, not ranked.
Tier = enrichment cadence driven by rank (top 10 = daily / next 40 = weekly / rest = monthly). Cold-start names haven't completed the minimum enrichment runs yet.
Click any row to expand the score breakdown. EXIT NOW = inside the 2–4 week pre-binary exit window.
Connect with your token to load the leaderboard.
Add Candidate
Identification
Don't know the ticker? Search by company name:
Hard Gates
Any UNKNOWN gate input → candidate PARKED (not rejected). All must pass for V to be computed.