Applied Post-Training AI Research Scientist

Baseten; San Francisco (CA); On-site; USD 180,000 - 250,000; Full time

Benefits offered by this job

Competitive compensation

Equity

Medical, dental, vision

Flex PTO

Parental leave

401(k)

Learning opportunities

Job summary

Baseten seeks a senior ML researcher to pursue open problems at the intersection of post-training methodology and scalable inference. You will translate findings into Baseten's production platform, collaborating with research engineering to ship AI products for customers.

A significant portion of your time will explore model learning, alignment, and architectural efficiency, with the remainder solving concrete platform challenges for customers using billions of tokens per second.

Qualifications

• Master’s or PhD in machine learning with first‑author publications at top venues.
• Ability to move from theory through implementation to empirical results.
• Judgment about problem selection to impact metrics and systems.
• Experience operating in a startup environment with product relevance.

Responsibilities

• Define and pursue a research agenda spanning foundational and applied work linked to Baseten's platform.
• Design rigorous experiments at meaningful scale (multi-node, trillion-parameter models).
• Collaborate with customers to translate requirements into research problems.
• Publish at top venues (NeurIPS, ICML, ICLR) and build Baseten's research presence.
• Bridge research findings with inference production systems alongside performance teams.
• Mentor junior researchers and shape the research organization’s direction.

Skills

Research depth

First-author publications

Open-source contributions

Cross-disciplinary

Judgment in problem selection

Education

Master’s or PhD in ML

Job description

Baseten seeks a senior ML researcher to pursue open problems at the intersection of post-training methodology and scalable inference. You will translate findings into Baseten's production platform, collaborating with research engineering to ship AI products for customers.

A significant portion of your time will explore model learning, alignment, and architectural efficiency, with the remainder solving concrete platform challenges for customers using billions of tokens per second.