Research, Evals

This full-time position at Exa in San Francisco focuses on designing and constructing the company’s evaluation infrastructure for its search engine. The day-to-day work involves investigating how to measure search quality in an era of large language models and building a comprehensive suite of tests. You will be responsible for creating creative measurement strategies and inspecting edge cases to define what constitutes effective search. This role directly influences the research team’s priorities and shapes the overall direction of the company’s products. Candidates need hands-on machine learning experience, specifically in training, fine-tuning, or evaluating models, with a bonus for background in embeddings or LLMs. Strong engineering fundamentals are required, including the ability to build reliable systems using Python, Rust, distributed pipelines, and GPU or cluster jobs. The advert does not specify any particular language proficiency requirements beyond the technical skills listed. It also does not detail specific office, hybrid, or remote working arrangements, only noting the location as San Francisco. This opportunity suits individuals who enjoy diving into data and are interested in defining the future of information retrieval. The company is an applied AI lab that has raised significant funding and powers search for major clients, aiming to build systems that filter global knowledge effectively.