2 days ago
South San Francisco, CA, USASenior / Staff+
H1B sponsor
Base Salary
$168k - $374k/yr
Responsibilities
- Own and evolve the backend architecture of the agentic molecule design platform, including service boundaries, APIs, persistence, workflow state, asynchronous execution, and integration patterns.
- Scale stateless services for projected users through horizontal replication, capacity testing, and reliable handling of sessions, workers, and database connections.
- Design resilient long-running scientific and agentic workflows with retries, checkpointing, idempotency, failure recovery, and human-in-the-loop interactions.
- Build backend services integrating AI agents, scientific tools, machine learning models, databases, and internal services.
- Evaluate AI agents and safely deploy selected agents using sandboxing, scoped permissions, monitoring, validation, and rollback.
- Establish engineering practices for testing, observability, reliability, performance, deployment, and operational ownership.
- Lead major platform workstreams and coordinate dependencies, risks, and stakeholder alignment across ML, data, and scientific teams.
- At the Principal level, own platform-wide technical direction and outcomes, lead across team boundaries, set reusable engineering standards, and mentor senior engineers.
Requirements
- Significant experience building and owning production backend, platform, or distributed systems.
- BS, MS, or PhD in Computer Science, Software Engineering, or a related field, or equivalent practical experience.
- Strong Python engineering skills and deep fundamentals in system design, APIs, distributed systems, data modeling, testing, and maintainable architecture.
- Experience designing asynchronous workflows with persistent state, failure handling, and clear service boundaries.
- Experience migrating a production system to a new framework or backend while minimizing downtime.
- Experience horizontally scaling services and handling failure modes involving in-memory state, sticky sessions, background workers, and database connections.
- Understanding of production AI-enabled applications, including observability, evaluation, deployment, reliability, and iteration.
- Ability to turn ambiguous product or scientific requirements into technical architectures and implementation plans.
- Experience leading complex workstreams and coaching engineers.
- For the Principal level, typically BS plus 7–12 years, MS plus 5–10 years, or PhD plus 2–7 years of relevant experience, with ownership of architecture and outcomes for complex multi-team platforms.
- For the Principal level, strong knowledge of LLMs, agentic frameworks, evaluation, model validation, autonomous-agent guardrails, cross-team influence, and senior-engineer mentorship.
- Preferred experience with agent orchestration, tool calling, state management, human-in-the-loop workflows, retrieval, tracing, Model Context Protocol, model serving, scientific computing, workflow engines, event-driven architectures, queues, distributed job execution, data platforms, cloud infrastructure, containers, Kubernetes, Helm, load testing, infrastructure as code, production observability, computational biology, bioinformatics, cheminformatics, computational chemistry, protein engineering, drug discovery, or safe execution of untrusted or model-generated code.
Benefits
- Benefits are available as detailed at the provided link.
- A discretionary annual bonus may be available based on individual and company performance.
- Relocation benefits are not available for this job posting.
- The stated California salary ranges are $168,100–$312,300 for Senior AI Engineer and $201,300–$373,800 for Principal AI Engineer.
Tech Stack
Categories
About Roche
Roche is a Swiss healthcare company that develops and sells prescription pharmaceuticals and in vitro diagnostic systems for hospitals, laboratories, and clinicians. Its business spans two core divisions—Pharmaceuticals and Diagnostics—with strengths in oncology, immunology, neuroscience, infectious diseases, and diabetes care. Founded in 1896 and headquartered in Basel, Switzerland, Roche also advances personalized healthcare by pairing therapies with companion diagnostics.
