Job Description
Help train large-language models (LLMs) to write production-grade code across a wide range of programming languages. The role involves comparing and ranking multiple code snippets, explaining which is best and why, also, repairing & refactoring AI-generated code for correctness, efficiency, and style. Feedback (ratings, edits, test results) is injected into the RLHF pipeline and keeping it running smoothly. The end result is the model learns to propose, critique, and improve code.
Engineers will need 4+ years of professional software engineering experience in Python. Strong code-review instincts are needed in order to spot logic errors, performance traps, and security issues quickly. A strong attention to detail and excellent written communication skills are a must. Enjoying reading documentation and language specs and thriving in an asynchronous, low-oversight environment is also important.