Sepal AI

Compensation

Full Description

Sepal is a data research company on a mission to advance human knowledge and capabilities through safe AI. We partner with the world’s leading AI labs and enterprises to help their models get better at the tasks people actually want them to do. In this role, you will work with top researchers, domain experts, and a team of senior engineers on open-ended projects. You will design and implement datasets, systems, and workflows from scratch and help push the boundaries of what LLMs can do. RL Data Design & Creation Build high-quality evals, training tasks, and verifiers across multiple domains. Build RL environments with MCP tools and computer-use interfaces. Build “recipes” and automations that make it easy for domain experts to create evals. Pair with LLM researchers to design training strategy and data. Open-Ended Research Propose and design novel post-training datasets and benchmarks for SOTA models. Prototype your ideas and POC sample datasets. Platform Engineering Shape the direction and feature set for data annotation platforms. Implement distributed systems for running agentic tasks. Skills Required Strong Docker and container-engineering skill. Experience with DevOps, CI/CD, and AWS. Highly proficient in Python and/or Typescript. Strong interpersonal skills and the ability to work with multiple teams. Good to have: experience with RL, SFT, and other post-training techniques. Good to have: in-depth knowledge on any specialized domains or sciences.