Research Scientist Intern: Frontier AI & RL Environments

Collinear AI, Inc.; Sunnyvale, Northern (CA, KY); Hybrid; USD 180,000 - 250,000; Full time

Job summary

Collinear AI, Inc. seeks a Member of Technical Staff (MTS - Research) to build the data engine for frontier AI in Sunnyvale, CA.

You will bridge frontier research and production engineering, developing high-fidelity environments and evaluation stacks used by leading AI labs to stress-test advanced agents. You will iterate on novel RL approaches and translate them into scalable infrastructure, contributing to reliability, safety, and real-world model metrics.

Qualifications

• Bachelor’s, Master’s or PhD in a technical field (CS, Math, Physics, etc.) or demonstrated open-source/industry work.
• Strong software engineering foundation; robust, scalable infrastructure; Python-friendly CLI-first development.
• Principled understanding of foundation models, how they are constructed, evaluated, and optimized.
• Experience conducting research or technical experiments with reproducible, data-driven results.

Responsibilities

• Build Agentic Environments: design and implement ultra-realistic, long-horizon simulation environments for agents learning to navigate ambiguity and context.
• Programmatic Verification: develop policy-aware judges and evaluations that measure capability and safety beyond benchmarks.
• Close the Loop: design and run post-training assessments (CPT, SFT, RL) to attain frontier performance on open-source models.
• Rapid Iteration: debug and iterate across the ML stack from infrastructure to model behavior, with a CLI-first focus.
• Collaborate: work daily with founders and research staff to shape the roadmap and push AI reliability.”

Skills

ML Fluency

Research

Python CLI

Engineering Rigor

Education

Bachelor’s/Master’s/PhD in CS/Math/Physics

Job description

Collinear AI, Inc. seeks a Member of Technical Staff (MTS - Research) to build the data engine for frontier AI in Sunnyvale, CA.

You will bridge frontier research and production engineering, developing high-fidelity environments and evaluation stacks used by leading AI labs to stress-test advanced agents. You will iterate on novel RL approaches and translate them into scalable infrastructure, contributing to reliability, safety, and real-world model metrics.