About the role#
The NeoSigma Research Fellowship is a 10-14 week, in-person, paid program for researchers and builders to work on open, real-world AI challenges. Fellows collaborate with the core research team on problems at the frontier of self-improving AI systems. Projects are open-ended, allowing you to define a research direction that aligns with your strengths, interests, and long-term goals. We encourage fellows to contribute to the broader research community by publishing findings through papers, technical blogs, and open research artifacts.
What you'll do#
- Work on open-ended problems in self-improving AI, including agentic systems, auto-evals, verifier research, and benchmarks.
- Define your own research direction based on your strengths and interests.
- Run experiments, iterate quickly, and push ideas from hypothesis to working system.
- Publish findings via papers, technical blogs, and open research artifacts.
- Collaborate daily with the research team through mentorship and brainstorming sessions.
What you'll need#
- Currently pursuing or holding a Master's or PhD degree in Computer Science, Machine Learning, Mathematics, Physics, Electrical Engineering, or a related field.
- Hands-on experience building LLM-powered or agentic systems in production.
- Strong research taste in evals, post-training, or agent architecture.
- Ability to own problems end-to-end, from first idea to shipped artifact.
- Comfort working directly with a small, fast-moving team.
- High attention to detail across research rigor, code quality, and writing.
Location & details#
- Location: San Francisco, California.
- Term: Summer 2026.
- Modality: On-site, full-time.
- Sponsorship: Visa sponsorship is available.
