Applied AI Researcher: Build Trusted Multi‑Step Agents

Bolter; United Kingdom; On-site; GBP 70,000 - 110,000; Full time

Job summary

Bolter is seeking an AI Researcher to design and run experiments that test how our agents behave, retain context, and complete multi-step tasks reliably, with a focus on trusted outcomes. You’ll work directly with the GM and the product engineer in a lean team, shaping the roadmap and influencing how AI agents are trusted and used in real work.

You’ll build and own the evaluation layer, translate research into actionable recommendations for the engineering team, and prototype ideas into working

Qualifications

• Applied AI/ML research track record in LLM apps, agents, or production reliability.
• Strong experimental design and rigorous analysis.
• Strong engineering fundamentals; able to prototype experiments.
• Deep familiarity with LLMs, tool use, context management, and agentic failure modes.
• Strong written communication; able to translate research into action.
• Rigorous yet pragmatic; know when to act on findings.
• Comfort with ambiguity; solving open-ended problems.
• Ability to leverage AI agents as a force multiplier.
• Ship-focused; craft and delivery matter.

Responsibilities

• Design and run experiments to test agent behavior and reliability.
• Build and own the evaluation layer; design meaningful evals.
• Research the frontier of LLM agents and translate to product.
• Turn findings into actionable engineering recommendations.
• Prototype research ideas into working prototypes and hand off.
• Shape the research roadmap as Bolter grows.

Skills

Applied AI/ML research

Experimental design

Engineering fundamentals

LLMs & tool use

Technical communication

Ambiguity tolerance

Agentic systems

Product-minded research

Shipping-focused

Job description

Bolter is seeking an AI Researcher to design and run experiments that test how our agents behave, retain context, and complete multi-step tasks reliably, with a focus on trusted outcomes. You’ll work directly with the GM and the product engineer in a lean team, shaping the roadmap and influencing how AI agents are trusted and used in real work.

You’ll build and own the evaluation layer, translate research into actionable recommendations for the engineering team, and prototype ideas into working