ML Research Scientist, Environments & Harnesses
Core
Designing synthetic RL environments, multi-agent harnesses, and end-to-end pipelines to enable AI-driven scientific discovery.
Role type
Senior ML Research Scientist (RL Environments & Multi-Agent Systems)
Builds
Synthetic RL environments, multi-agent harnesses, and scalable training pipelines for scientific discovery.
Domain
AI for Science (RL, Multi-Agent Systems, Scientific Workflows)
Deliverable
production ML models
Required skills
Reinforcement Learning, Multi-Agent Systems, Python, Prototyping, Software Engineering, Experiment Design, Reward Shaping, Hierarchical Planning
Preferred skills
Publications (NeurIPS/ICML/ICLR), Domain Science (Biology/Chemistry/Physics), Distributed Compute (Ray/Modal), LLM Inference (vLLM), Workflow Orchestration (Flyte/Airflow)
Responsibilities
Design and prototype new harness capabilities (context compaction, tool interfaces, long-horizon planning, multi-agent swarms). Invent new types of RL environments by transforming scientific workflows and simulators into training signals. Build end-to-end pipelines for synthetic environment generation from raw assets. Prototype and validate environment designs including reward shaping and integration with training loops. Push boundaries of multi-step reasoning for chained tasks and complex evaluation scenarios.
Seniority
Senior, hands-on IC
