AI Agent Engineer, Evaluation and Reliability (Semiconductor AI)

A fast-growing AI company that builds software agents for semiconductor and hardware engineering is hiring engineers to make those agents stronger, more dependable and cheaper to run. The role blends talking with customers, designing evaluations, applied research and production engineering, alongside software engineers, researchers and hardware specialists.

Your focus

Identify where agents fail, then build and execute evals that improve quality, reliability and efficiency

Create new agent capabilities: interfaces to tools, handling of context, memory, and orchestration across several steps or several agents

Examine run traces and test outcomes to pinpoint slowdowns, and distinguish true progress from benchmark noise

Talk with customers and subject specialists about their needs, then ship and maintain solutions through to completion

Turn experiments that work into solid, maintainable software in production

Profile

Five to fifteen years of software engineering, including work on large-scale agent systems

Strong production skills in Python and TypeScript

Working knowledge of current language models, agent frameworks and typical ML workflows

Experience with LLM agent orchestration, tools, memory and tracing

Communicates well, shows sound judgment about customers and can isolate the core of vague technical problems

Self-directed with strong ownership and product sense

Bonus: interest in chip design or systems that span hardware and software

On-site in the South Bay Area (Silicon Valley). Not able to offer visa sponsorship or transfers.