Staff Engineer, RL Data & Environments

Mirendil Inc.; San Francisco (CA); Hybrid; USD 300,000 - 400,000; Full time

Benefits offered by this job

Equity grant

Competitive benefits

Job summary

Mirendil is seeking a Research Engineer to own the data systems and execution environments that power reinforcement learning. You will design and build end-to-end pipelines and infrastructure to ensure high‑quality data and robust environments for training complex tasks.

The role involves collaborating across teams to push axes of improvement in model behavior and to scale sandboxed environments for multi-agent workloads.

Responsibilities

• Build and automate data collection pipelines for long-horizon RL tasks.
• Prevent reward hacking with robust data and environment safeguards.
• Create scalable sandboxed execution environments for multi-agent tasks.
• Design metrics to estimate how training environments influence production model behavior.
• Collaborate across teams to improve training environments and push for better model behavior.

Job description

Mirendil is seeking a Research Engineer to own the data systems and execution environments that power reinforcement learning. You will design and build end-to-end pipelines and infrastructure to ensure high‑quality data and robust environments for training complex tasks.

The role involves collaborating across teams to push axes of improvement in model behavior and to scale sandboxed environments for multi-agent workloads.