AI Researcher

About the role

As an AI Researcher at Bolter, you will design and run experiments to make our autonomous agents trustworthy, reliable and capable of completing multi‑step tasks. This applied research role directly influences product decisions and works closely with the GM and product engineer in a lean, fast‑moving team.

Key responsibilities

• Design and execute experiments evaluating agent reliability, context retention and multi‑step task completion.
• Build and own the evaluation framework that measures actual work outcomes rather than plausibility.
• Stay up‑to‑date with state‑of‑the‑art LLM agents, tool use and reliability research, translating insights into product road‑maps.
• Convert research findings into clear, actionable recommendations for engineering implementation.
• Prototype promising ideas into working prototypes and hand them off for production.
• Shape the research roadmap as the company scales.

Required profile

• Proven track record in applied AI/ML research, preferably with LLM‑based applications, agentic systems or reliability evaluation.
• Strong experimental design and analytical skills; ability to test hypotheses rigorously.
• Solid engineering fundamentals enabling you to prototype experiments yourself.
• Excellent written communication to convey research outcomes to product and engineering teams.
• Pragmatic and rigorous mindset, comfortable with ambiguity and deciding when evidence is sufficient for action.

Required skills

• Experience with large language models (LLMs) and tool‑use integration.
• Proficiency in building evaluation pipelines for AI agents.
• Programming skills in Python (or comparable language) for prototyping experiments.