Machine Learning Engineering Intern, Evals/Post-training

Winter 2026 (January - April) Internship - full-time

Hybrid (Palo Alto, CA)

Mission

We’re a small, fast team behind OpenBench (open, reproducible LLM evals). We turn model behavior into measurable progress, then upstream it. You’ll work alongside people, not for people: low ceremony, quick feedback, lots of ownership. You won’t be siloed; you’ll jump across evals, post-training, infra, and (when useful) product/GTM.

Responsibilities & Opportunities In This Role

• Build and reimplement evals (accuracy, robustness, safety, latency) end-to-end.
• Run tight SFT/DPO/RLHF-style loops; track deltas and ship models for customers.
• Red-team models; turn quirks into metrics and provide feedback to the inference team
• Own scoped projects: design → implement → document → upstream.
• Write research papers on evals you build.
• Pitch improvements across the company when you see them, then ship.
  
  

Ideal Candidates Have/are

• Founding Engineer (grinder)
• You unblock yourself, learn fast, and ship relentlessly - scrappy first, then clean and reproducible.
• Signals: productionized side projects, CI’d repos, tools other people actually use.
• Researcher (loves data and pushing the frontier)
• You reason clearly about eval design, failure modes, and data quality; you run ablations and write tight analyses.
• Signals: careful experiments, thoughtful write-ups, PRs to open-source projects.
• Must-haves
• Agentic, kind, gritty.
• Hands-on with evals, post-training, or applied AI (not just theory).
• Comfort getting a bit hacky while keeping results reproducible.
  

Why Join Us

• Purposeful Hiring: You’re not here by accident, and neither is anyone else. Every teammate is handpicked with intention because who we build with matters.
• Builders Wanted: You’re not just riding the rocket ship, you’re building it. Your work directly shapes the trajectory of our company.
• Mission-Driven Work: We’re here to make a real impact. Our mission fuels everything we do.
• Tackling Hard Problems: If easy isn’t your thing, you’re in the right place. We solve some of the most complex and exciting challenges in our space.
• Excellence Is The Standard: High performance isn’t just encouraged, it’s the baseline. And it’s contagious.
  
  

If this sounds like you, we’d love to hear from you!

Compensation: The US pay range for our technical internships is $30-$50 / per hour. The pay range for our non-technical internships is $30-$40 / per hour. Compensation is determined by your location, skills, qualifications, experience and internal benchmarks. This range is specific to roles in the United States, compensation for candidates outside the USA will be dependent on the local market.

This position may require access to technology and/or information subject to U.S. export control laws and regulations, including the Export Administration Regulations (EAR). To comply with these requirements, candidates for this role must meet certain citizenship or residency criteria. Specifically, they must qualify as U.S. Persons for export control purposes (i.e., U.S. citizen, U.S. lawful permanent resident (Green Card holder), or a protected individual under 8 U.S.C.

• 1324b(a)(3) such as a refugee or asylee), or otherwise be eligible for an applicable export license.