Principal Scientist, Translational AI Foundations

Bristol Myers Squibb • Needham, MA

6d • $166,770 - $202,086

About The Position

The Principal Scientist will own the shared software platform for Translational AI Foundations (TAIF), a small team of computational biologists and AI engineers working alongside translational research groups. TAIF converts biomarker analyses performed manually by scientists into validated, reusable AI agents and workflows, then transfers them to the research groups that will use and maintain them. This position builds shared tools and foundations rather than owning the scientific work of those groups.

The role focuses on creating engineering-standard tooling, operating the team's ontology and knowledge graph as shared services, and building versioned connections to governed data. It also deploys validated tools as reliable services with CI/CD, monitoring, and rollback plans. The initial scientific focus is circulating tumor DNA (ctDNA), including its data connection and reference dataset, followed by a ctDNA agent co-developed with the team's assay scientist after the first spatial agents. Deployment refers to engineering deployment; the team lead guides broader adoption, with this role providing hands-on support.

The successful candidate will build on established enterprise data platforms in partnership with data, IT, and knowledge-management teams. The role is primarily devoted to the shared platform and knowledge systems, with the remainder focused on deployment, hand-off, measurement, and direct partnership with translational research groups. Mature components will ultimately be operated by established platform, research engineering, and IT teams.

Requirements

• Ph.D. in computational biology, bioinformatics, computer science, data engineering, or a related quantitative field and at least 4 years of relevant academic and/or industry experience; or a master's degree and at least 6 years; or a bachelor's degree and at least 8 years.
• Preferred: Record of shipping production data systems or services used by other teams, such as Python services, SDKs, or pipelines, with tests, version control, and code review.
• Preferred: Hands-on cloud engineering on AWS or an equivalent provider, including storage and compute, containers, infrastructure as code, CI/CD, and monitoring and troubleshooting of running services.
• Preferred: Experience designing governed FAIR data foundations with data owners, including schema and entity modeling, metadata harmonization, data-version and lineage discipline, access control, and retention.
• Preferred: Working familiarity with translational biomarker data, including sequencing-based assays, liquid biopsy, clinical and sample metadata, and GxP-relevant traceability expectations such as audit trails, versioning, validation evidence, and human-review records.
• Preferred: Experience building LLM-based or agentic automation, or building the data connectors, benchmark datasets, and evaluation on which such systems depend, with the engineering judgment to operate them as maintained services.
• Preferred: Production experience with agentic systems, including tool servers, retrieval and grounding, state and memory, evaluation harnesses, regression testing, telemetry, and multiple model providers behind one gateway.
• Preferred: Knowledge graph and ontology engineering experience, including graph databases, controlled terminologies, data dictionaries and standards, schema validation, and provenance models.
• Preferred: Experience with container orchestration and platform engineering at scale, including Kubernetes, EKS or ECS, Terraform, event-driven architectures, workflow managers, and secondary-analysis platforms.
• Preferred: Experience with governed data products such as Databricks, data-version discipline, observability, and cloud AI-service cost management.
• Preferred: Regulated delivery experience, such as companion-diagnostic or IVD validation, quality-management-system requirements, CAPA resolution, or traceable evidence for IND or marketing submissions.
• Preferred: Experience in a platform or enablement team serving scientists directly and measuring success by adoption, including workshops, office hours, or governed SDKs used by bench scientists.
• Preferred: Front-end experience such as Streamlit, JavaScript, or TypeScript.
• Preferred: Experience leading engineers or cross-functional delivery involving assay development, biostatistics, clinical, IT, legal, and regulatory stakeholders.
• Preferred: Prior experience in translational oncology or early clinical development at a biopharmaceutical or diagnostics company.

Nice To Haves

• A record of shipping production data systems or services used by other teams, such as tested Python services, SDKs, or pipelines.
• Hands-on cloud engineering on AWS or an equivalent provider, including storage and compute, Docker, infrastructure as code, CI/CD, monitoring, and troubleshooting.
• Experience designing governed FAIR data foundations, including schema and entity modeling, metadata harmonization, lineage, data versioning, access control, and retention.
• Working familiarity with translational biomarker data and GxP-relevant traceability expectations.
• Experience building LLM-based or agentic automation, or the data connectors, benchmark datasets, and evaluation systems on which such automation depends.
• Production experience with agentic systems, tool servers, retrieval and grounding, state and memory, evaluation, regression testing, telemetry, and model-provider gateways.
• Knowledge graph and ontology engineering, including graph databases, controlled terminologies, standards, schema validation, and provenance models.
• Container orchestration and platform engineering at scale, including Kubernetes, EKS or ECS, Terraform, event-driven architectures, workflow managers, and secondary-analysis platforms.
• Experience with governed data products such as Databricks, data-version discipline, observability, and cloud AI-service cost management.
• Regulated delivery experience, including companion-diagnostic or IVD validation, quality-management systems, CAPA, or traceable evidence for submissions.
• Experience in a scientist-facing platform or enablement team focused on adoption, including workshops, office hours, or governed SDKs.
• Front-end experience such as Streamlit, JavaScript, or TypeScript.
• Experience leading engineers or cross-functional delivery across scientific, clinical, IT, legal, and regulatory stakeholders.
• Prior experience in translational oncology or early clinical development at a biopharmaceutical or diagnostics company.

Responsibilities

• Build and maintain shared platform capabilities, including provenance tracking for data, code, model, and prompt versions; evaluation tooling; access control; logging; and usage telemetry, in accordance with enterprise security and release standards.
• Turn engineering standards into tooling by implementing tool templates, automated test and validation pipelines, packaging for AI-agent use (such as Model Context Protocol tool servers), and a registry of tools, owners, versions, lifecycle stages, and users.
• Own the ontology and knowledge graph service that holds canonical biomarker definitions, beginning with a biomarker-trial ontology covering subjects, cohorts, visits, specimens, assays, endpoints, and cutoffs, and connect it to the enterprise data catalog.
• Build and maintain approved, versioned connections to governed data products and biomarker data stores, record the data version on every read, map clinical studies into a common data model, and coordinate data logistics and governance with data owners, starting with ctDNA.
• Containerize and deploy each validated tool and service using Docker and an appropriate deployment platform, with CI/CD, monitoring, and a rollback plan; own the release checklist with platform teams.
• Package components for their long-term owning teams with documentation, tests, versioning, and registry entries linked to test cases so they can be taken over unchanged.
• Instrument tools for usage reporting, including users, teams, elapsed and review time, test results, and use after hand-off; report usage and impact metrics.
• Work with AI contacts in translational research groups by running hands-on tool-building sessions and reviewing their tools for engineering-standard compliance and deployability.
• Set engineering practices for the team and AI contacts, including code review, testing, versioning, and release discipline; mentor junior team members and document the shared platform so its owning team can operate it independently.
• Co-build the ctDNA agent with the team's assay scientist after the first spatial agents, using the shared platform and the assay scientist's scientific test cases and pass criteria.

Benefits

• Comprehensive benefit plans, including wellbeing support, retirement and financial protection benefits, and medical, dental, vision, life, and disability insurance offerings.
• Flexible Time Off without a set accrual limit for eligible U.S.-based exempt employees, subject to manager approval.
• 11 paid company holidays each year for eligible U.S.-based exempt employees.
• Depending on eligibility, paid sick leave, up to two paid volunteer days per year, summer hours flexibility, and leaves of absence for medical, personal, parental, caregiver, bereavement, or military needs.
• Eligible employees receive an annual Global Shutdown between Christmas Day and New Year's Day.
• Potential eligibility for discretionary incentive cash and stock opportunities.

What This Job Offers

• Job Type: Full-time
• Career Level: Principal
• Education Level: Ph.D. or professional degree