Machine Learning Researcher

About Pinetree

Pinetree is a frontier lab focused on transforming general intelligence into reliable labor. We build autonomous, long-horizon Computer Use Agents, which we believe is the first step towards achieving AGI.

Role Overview

We are looking for a Machine Learning Researcher to work on core research problems in agentic systems, multimodal models, and learning under real-world constraints. This role is research-driven but tightly coupled to deployment. You will design, prototype, and evaluate novel ML methods that enable reliable computer use, long-horizon reasoning, and robust interaction with complex interfaces.

This is an ideal role for someone who enjoys working on open research questions but wants their work to ship and matter.

Responsibilities

• Conduct research on agentic ML systems, including planning, reasoning, tool use, and long-horizon execution
• Design and evaluate learning methods for computer use, including vision-language models, multimodal perception, and UI understanding
• Develop benchmarks, datasets, and evaluation frameworks for agent reliability, robustness, and safety
• Experiment with fine-tuning, distillation, and reinforcement learning for real-world task execution
• Analyze agent failures and propose principled fixes grounded in theory and empirical results
• Collaborate closely with engineers to translate research ideas into production systems
• Author internal research reports and, when appropriate, external publications

Required Qualifications

• Strong background in machine learning, AI, or a related field
• Ability to design rigorous experiments and analyze results critically
• Comfortable working in fast-moving, ambiguous problem spaces

Preferred Qualifications

• PhD or industry research experience in agents, robotics, HCI, or applied ML
• Familiarity with browser automation, web agents, or UI interaction models
• Publications in NeurIPS / ICLR / ICML / ACL / EMNLP or other respected conferences or journals

What We Offer

• Opportunity to work on foundational problems in real-world AI agents
• Direct ownership over research direction and impact
• Close collaboration with a highly technical founding team
• A culture that values rigor, clarity, and shipping real systems