Intern, AI Agent Evaluations

What you'd do

• Break an AI agent's job (for example, outbound sales) into steps that can be tested. Write the tasks, test data, and scoring rules for those steps.
• Review the scores our AI judge and human reviewers give.
• Support scaling system to engage with hundreds of vendors.
• Prepare the results packet we share with each vendor before we publish, and keep the vendor pipeline current every week.
• Track who else evaluates AI agents (AI labs, benchmark groups, evaluation startups, analyst firms) and what they publish.

What they want

• Currently enrolled in a MBA.
• Have built with or used AI tools such as LLM APIs, chat assistants, or agents, and can explain how they work and where they fail.
• Write clearly and briefly.
• Comfortable talking with startup founders and product leaders. Able to hold your view when someone pushes back, and to change it when the facts say so.
• Able to make progress on problems that have no set process yet.
• Based in the San Francisco Bay Area.
• Available 10 to 20 hours a week during the school year.
• The internship starts October or November 2026.
• You have built an AI agent, an evaluation, or a benchmark, even as a class or side project.
• You have worked at or with a B2B software company in any role: startup, venture, sales, or product.