Eval Researcher — Speech & Text Models San Francisco - Onsite We’re working with a frontier AI research lab building real-time voice agents that listen, respond and interact naturally alongside people. Their first application is AI gaming companions that understand live gameplay, handle interruptions and participate in multiplayer conversations. The team also produces research-grade voice data for frontier model developers, with research expanding into robotics and physical AI. They’re hiring an Eval Researcher who has: • Created evals for speech and/or text models used in production. • Used those evals to improve the quality of the product experience. • Kept up to date with the latest evaluation methods and research, with examples from commercial projects, academia or personal research.