Build and operate infrastructure for evaluating coding agents and language models. Design end-to-end features spanning data models, APIs, background workers, and UI; run sandboxed trials; score and aggregate results; improve benchmark quality checks; integrate models and agents; and maintain performance, reliability, concurrency, migrations, tests, and code quality.