AI Researcher - Bolter

Role overview

An applied AI researcher is needed to make autonomous agents that carry out real work trustworthy, capable, and useful. The work sits squarely between research and product — designing experiments, running evaluations, and turning findings into shipped decisions rather than papers. The role joins a lean founding team working directly with the GM and a product engineer, supported by specialized AI agents, during the platform's validation stage.

Responsibilities

- Design and run experiments that test how the platform's agents behave — reliability, context retention, multi-step task completion, and failure modes. - Build and own the evaluation layer, designing evals that measure whether agents actually complete work, not just whether they sound plausible. - Track the state of the art in LLM agents, tool use, and reliability, and translate those findings into recommendations for what to build next. - Produce clear, actionable recommendations the engineering team can ship directly into the product. - Prototype promising research ideas into working features and hand off what proves useful to engineering. - Shape the research roadmap as the platform grows from the validation stage into broader rollout.

Requirements

- A track record of applied AI or machine learning research, ideally in LLM application design, agentic systems, evaluations, or production AI reliability. - Strong experimental design and analysis skills, with the ability to test a hypothesis rigorously and read results honestly. - Solid engineering fundamentals, with the ability to prototype your own experiments rather than only directing others. - Deep familiarity with large language models, tool use, context management, and the failure modes of agentic systems. - Strong written communication, since research that is not understood and acted on does not help. - Comfort with ambiguity, pragmatism about evidence, and the ability to use AI agents as a force multiplier on a small team.

Nice to have

- Experience with LLM evaluations, hallucination mitigation, or production AI reliability at scale. - Experience building or evaluating agentic systems, tool use, or autonomous workflows. - A public track record such as papers, open source, blog posts, or side projects. - Background in product-led research where the output is a shipped feature rather than only a finding.