Overview
We’re looking for experienced drug discovery scientists to help train AI models. You’ll design realistic discovery and preclinical tasks from your own practice, run them through frontier AI systems, and grade what comes back against a professional standard.
The models can talk fluently about drug discovery. What they can’t yet do reliably is the real work: SAR analysis, screening triage, DMPK and PK/PD interpretation, or a candidate-selection call. Your judgment moving programs from target to IND becomes the benchmark those models are measured against.
What you’ll actually do
- Design realistic discovery tasks from your own workflows: the scenario, the prompt, and the files an agent would need (assay result tables, SAR spreadsheets, DMPK summaries, study reports, program-review decks).
- Run tasks through frontier AI agents and grade the deliverable (a decision memo, screening workbook, or tox summary) against the standard you’d hold a colleague to.
- Write clear grading rubrics (the right compounds advanced, the right liabilities flagged, the right calculations) and explain why a response passes or fails.
- Flag concrete failures with evidence: misread assay data, SAR conclusions the data don’t support, PK parameters misinterpreted, unit and scaling errors, or missed safety liabilities.
Problems draw on whatever you know best
- Medicinal chemistry and SAR / lead optimization.
- Computational chemistry and CADD.
- Assay development and screening.
- DMPK, PK/PD, ADME, and preclinical safety.
- Translational and biomarker science, biologics and advanced modalities, and CMC / process development.