Remote Coding RL Engineer - Benchmarks & Reproducible Tasks

Clera Labs, Inc.; San Francisco (CA); Remote; USD 180,000 - 260,000; Full time

Job summary

Clera Labs, Inc. is seeking a Coding RL Engineer to join the reinforcement learning team. You will turn real-world software work into reproducible benchmarks and build coding environments for training and evaluating advanced models.

You will convert production codebases into containerized environments, create tasks from real development activities, and help strengthen verifiers and benchmarking tools with research collaboration.

Qualifications

• 2+ years of professional software engineering delivering software systems.
• Strong Python skills and fluency in another language such as TypeScript, Go, Java, or Rust.
• Hands-on Docker experience and familiarity with reproducible build/test environments.
• Experience navigating unfamiliar, complex codebases and getting them building, running, and tested.
• Knowledge of coding model training and evaluation (SWE-bench-style tasks, code evaluations, RLVR).
• Experience creating coding tasks from real engineering work and hardening verifiers against flaky tests.

Responsibilities

• Convert licensed production codebases into reproducible, containerized coding environments.
• Create tasks from issues, PRs, commit histories, and incident records.
• Build verifiers with hidden tests and rubric- or model-based grading.
• Run model rollouts, review traces, and fix broken or ambiguous tasks.
• Contribute tasks and analysis to coding benchmarks and tooling for faster task creation.
• Collaborate with research teams to calibrate task difficulty and benchmark coverage.

Skills

Python

TypeScript/Go/Java/Rust

Docker

Codebase navigation

RL/ML knowledge

Tools

Docker

Job description

Clera Labs, Inc. is seeking a Coding RL Engineer to join the reinforcement learning team. You will turn real-world software work into reproducible benchmarks and build coding environments for training and evaluating advanced models.

You will convert production codebases into containerized environments, create tasks from real development activities, and help strengthen verifiers and benchmarking tools with research collaboration.