One-Liner
An AI-native platform that takes an entire chemical portfolio (500+ substances) and produces a REACH compliance triage report showing which substances need testing, which can use validated computational methods, and which need early reformulation — saving 60-80% of testing costs.
AI Thinking Process
REACH 2027 registration deadline: mid-size German chemical company needs to assess 150+ substances. Current cost: €5K-50K per substance via CRO, 3-6 months each. Total bill: €750K-€7.5M over 4-8 years
Applied Impossibility Negation: 'You can't assess chemical safety without animal testing' — increasingly FALSE. ECHA accepts QSAR models under REACH Annex XI, OECD-validated methods exist for multiple toxicological endpoints, EU committed to phasing out animal testing by 2035
Portfolio-level REACH triage: AI platform takes 500-substance portfolio, classifies each against ECHA requirements, matches endpoints to best validated computational methods, outputs testing strategy. Not 'AI fills registration forms' — strategic triage tool saving 60-80% testing costs
WHO: REACH compliance manager at mid-size German chemical company (200 employees, 500-substance portfolio). CURRENT: sends substances to CROs one by one at €5K-50K each, waits 3-6 months, tracks in spreadsheet. WHY-SURPRISED: ECHA has accepted QSAR data since 2013, AI models improved dramatically, EU banning animal testing — yet commercial tools are still desktop software from 2000s (Derek Nexus). No modern SaaS for portfolio-level screening exists
ChemCopilot (lifecycle management, not screening), Derek Nexus/Lhasa (desktop QSAR, single-substance), OECD QSAR Toolbox (free, expert-only, terrible UX), Schrödinger (drug discovery focus). No modern SaaS portfolio screening found
Kill reason: regulatory acceptance risk for novel AI models. Pivot: become modern SaaS frontend for existing validated tools (bundle Derek Nexus + QSAR Toolbox + VEGA with portfolio management). Addresses regulatory risk but makes product incremental — 'better UX for existing tools' (G020 incremental flag). Pivot failed — too incremental
48% conviction — 2 points below 50% painpoint minimum. Three kill factors: regulatory acceptance of novel AI toxicology models, entrenched Derek Nexus incumbent (15+ years regulatory trust), rare talent combination. Resurrection candidate: 2027 REACH deadline + regulatory acceptance timeline acceleration may push to 50%+
Reviewing REACH screening kill reasons: (1) regulatory acceptance, (2) Derek Nexus incumbent, (3) rare talent combination — classifying each as fundamental or positional
All three kill reasons are POSITIONAL: regulatory acceptance is evolving (ECHA 2025 'Key Areas' explicitly names computational methods, ICH M7() accepts AI for mutagenicity), Derek Nexus gap confirmed (desktop tool for toxicologists vs modern SaaS for compliance managers — different buyer), rare talent is team composition not market structural. Pivot to 'portfolio intelligence orchestration using validated QSAR methods' reaches 45% conviction. Flagged for founder review due to 2027 REACH deadline urgency. Falls short of 50% threshold.
Kill Reason
Below the 50% painpoint minimum threshold at 48% conviction. Regulatory acceptance risk for novel AI toxicology models, entrenched incumbent (Derek Nexus by Lhasa Limited, 15+ years standard), and rare talent combination required (toxicology PhD + software engineering) could not be overcome within threshold.
Risk Analysis
Risk analysis available for latest engine ideas.
What do you think?
Related ideas you can explore free:
killed: Open-source middleware (HAMi) already provides heterogeneous AI computing virtualization for free. Proprietary play is squeezed between free open-source and vertically integrated hardware vendor ecosystem.
killed: 5+ funded competitors including Cast AI ($1B valuation), OneChronos (backed by Nobel laureate), Akash Network (decentralized, 80% cheaper), Argentum AI (blockchain-settled). Market is claimed with massive capital.
killed: Template epidemic (G003) + industry-pain-form death pattern (G005) fire simultaneously. 13+ existing compliance tools. A prompt could do 80% of this.