SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Lila Sciences is seeking a Staff Research Engineer to bridge research and production in scientific computing and ML infrastructure. You will work with computational biophysics, chemistry, and ML scientists to transform prototype research tools into robust, scalable, and maintainable systems.
Key responsibilities include: taking research prototypes and scientific workflows and making them scalable, efficient, and production-ready; collaborating directly with researchers to convert exploratory work into agent-usable systems; building ML and physics infrastructure for model training, molecular simulation, data processing, and scientific workflows; ensuring reliability across multiple clusters and compute environments; optimizing GPU utilization, distributed execution, throughput, and fault tolerance; architecting larger systems with job orchestration, monitoring, artifact handling, and traceability; packaging scientific tools into reusable services and APIs; and establishing pragmatic engineering patterns for research teams.
Required qualifications: strong Python software engineering skills with experience in ML, scientific computing, or simulation codebases; hands-on experience building, scaling, or operating distributed systems for research or data-intensive workloads; practical knowledge of GPU computing, performance profiling, and distributed execution; experience with PyTorch, JAX, CUDA workflows, or similar frameworks; Linux, Docker, containers, and reproducible environments; orchestration tools like Kubernetes, Slurm, Ray, Flyte, or Argo; ability to improve research code architecture and scalability; strong debugging across code, environments, infrastructure, and clusters; and ability to work directly with researchers to convert ambiguous needs into robust solutions.
Bonus experience includes: computational chemistry, biophysics, molecular simulation, or drug discovery; cloud GPU infrastructure and multi-cluster execution; building tools for LLM agents or automated research; workflow observability and fault-tolerant scientific workloads; CI/testing/packaging practices; and comfort supporting fast-moving research teams without over-engineering.
