Convexia

Convexia is building an AI-maximalist pharma company. We replace months of manual drug asset diligence with automated asset sourcing, in silico scientific evaluation, commercial modeling, and clinical risk analysis. These agents feed into our proprietary PoS model to identify high potential drug assets.

We sell the platform to pharma BD teams, biotech startups, and rare disease clinician teams today. As we become system-integrated with customers, we will begin owning and advancing assets ourselves through key development milestones.

We build an agentic diligence system across Source → Diligence → Decide.

Sourcing agent: proprietary connectors that ingest and normalize across structured and unstructured sources, including non-English materials, tech transfer listings, and conference content.

Diligence agents:

• scientific in silico evaluation (binding and docking, ADME and PK, toxicity and safety),
• commercial diligence using proprietary claims data, and
• operational risk modeling using digital twin and Monte Carlo simulation (CMC, CRO fragility, site readiness).

Decision layer (PoS): a proprietary Probability of Success model that consumes outputs from all agents and can be customized to each client via reinforcement fine-tuning.