The opportunity
OpenAI’s Forward Deployed Engineering team partners with global pharma and biotech, CROs, and research institutions to deploy existing expertise across the R&D value chain to help customers design and ship production-grade AI systems. We operate at the intersection of customer…
What you'll do
Design and ship production systems around models, owning integrations, data: provenance, reliability, and on-call readiness across research, clinical, and operational workflows.
Lead discovery and scoping from pre-sales through post-sales, translating: ambiguous workflow needs into hypothesis-driven problem framing, system requirements, and an execution plan with measurable endpoints.
Define and enforce launch criteria for regulated contexts, including: validation evidence, audit readiness, outcome metrics, and drive delivery until we demonstrate sustained production impact.
Build in sensitive scientific data environments where auditability,: validation, and access controls shape architecture, operating procedures, and failure handling.
Run evaluation loops that measure model and system quality against: workflow-specific scientific benchmarks and use results to drive model and product changes.
Distill deployment learnings into hardened primitives, reference: architectures, validation templates, and benchmark harnesses that scale across regulated life sciences environments.
What they're looking for
- Bring 5+ years of software/ML engineering or technical deployment experience: with customer-facing ownership in biotech, pharma, clinical research, or scientific software; PhD, MS, or equivalent applied experience in a life sciences relevant field encouraged.
- Have owned customer GenAI deployments end-to-end from scoping through: production adoption, and improved them through evaluation design, error analysis, and iterative evidence generation that tightens acceptance criteria over time.
- Have delivered AI systems in trial design, regulatory writing, or scientific: operations where validation strategy, auditability, compliance constraints, and reviewer expectations shaped system design and rollout.
- Communicate clearly across scientific, clinical, model research, technical,: and executive audiences, translating technical tradeoffs into decision quality, risk posture, and measurable outcomes with credibility.
