The company is seeking experienced software engineers to audit and evaluate technical tasks used for training and evaluating AI systems. The role involves assessing task accuracy, solvability, and providing feedback on codebase integration and logic errors. Candidates should have deep technical knowledge in software engineering and SWE-Bench standards for this remote freelance contract.