FOXO
AI Intern
- Eligibility
- Not stated
- Batch
- Not stated
- Experience
- Not stated
- Stipend
- up to ₹60K/month
How well do you fit this role?
Your resume against this posting — what you have, what’s missing, and a short plan to close the gap.
PDF or text, up to 4 MB. It stays in this browser — nothing is saved to your account.
About this role
About the role: - FOXO is hiring 2 AI interns to work on engineering problems in preventive care. - The role is focused on building reliable AI systems rather than only writing prompts. - Interns will work directly with the team on identifying incorrect AI outputs, tracing their causes and preventing regressions. - The team operates with an AI-lab style approach and expects candidates to understand both product needs and implementation details. What you'll build: - Create evaluation harnesses combining deterministic checks with LLM-as-judge scoring. - Build evaluations that detect regressions when prompts or underlying models change. - Develop tracing systems that capture model steps, tool calls, latency and token costs. - Use traces to identify the exact point responsible for incorrect outputs. - Write test suites based on failures observed in real production systems. - Maintain versioned datasets with explicit schemas so dataset changes do not silently break evaluations. - Ship fixes using Claude Code and Cursor while ensuring changes are reviewed, tested and merged properly. - Turn production failures into reproducible tests and add resolved cases to regression suites. What we're looking for: - Candidates must be comfortable shipping code daily with Claude Code or Cursor. - You should understand the code being shipped rather than relying blindly on AI-generated implementation. - Solid Python skills are required, including APIs, async programming, JSON and pytest. - Applicants should be comfortable reading and debugging stack traces. - Experience using Git and working inside a real codebase is required. - Candidates should think about what the product and user need, not only execute narrowly defined tasks. Good to have: - Experience building evaluation systems and being able to justify the scoring methodology is valuable. - Experience with structured outputs, tool calling or agent frameworks is useful. - Ability to take an SOP and execute independently is considered a plus. Selection process: - The process starts with a 30-minute call with the FOXO team. - Candidates then complete one small task based on the team's actual work. - A hiring decision is expected within one week. Note: 2 openings. Email with subject: Your Name | AI Intern.