Senior / Principal AI Engineer

Senior / Principal AI Engineer (x2) - London (Hybrid) - Python, LangGraph, TypeScript, PostgreSQL, GCP

£90,000 - £150,000

I'm partnering with a well-funded AI scale-up that has just secured a major investment round and has the ambition to become a category-defining company. They're building an enterprise AI platform that helps large organisations get real, measurable value from AI, using intelligent agents to automate work at scale.

This is a rare chance to join at an early stage, with serious backing and a clear vision. You'll work directly with an experienced CTO from a blue-chip technology background, who is building the team in person in London. The business is scaling quickly across Tech, Product and Marketing, so you'll help shape a team of the best people in the market. The office is state of the art, in the heart of the City, and somewhere people want to come in to work.

You will work with

• Designing and building complex agent harnesses: runtime, orchestration, tool use, gates, stop conditions and recovery paths
• Building production LLM, RAG and agentic systems in Python and LangGraph
• Owning evaluation end to end: error analysis, trajectory-level testing, judge calibration, live quality monitoring and release gates
• Building secure integrations with business-critical systems, with human approval before consequential actions
• Developing sandboxed code execution and long-running, resumable background agent tasks
• Connecting cost, outputs and user feedback so customers can see the real value of their AI usage
• Partnering with the CTO on architecture, reviews, mentoring and engineering standards

Benefits

• Salary of £90,000 - £150,000 depending on experience, plus equity
• Join an early-stage company with major investor backing and the ambition to lead its market
• Direct access to an experienced CTO and real influence on technical direction
• Hybrid working: 3-4 days a week in a beautiful central London office
• Work on hard, meaningful problems rather than thin model integrations

Key experience

• Strong, hands-on experience building complex agent harnesses and evaluations (essential)
• Production experience with LLMs, RAG systems and AI agents
• Proficiency in Python and LangGraph, with solid engineering fundamentals (TypeScript, PostgreSQL, APIs, queues, cloud; GCP a plus)
• Experience evaluating agent trajectories, including tool selection, recovery, cost and reliability, and calibrating LLM judges
• Practical judgement on model quality, token budgets, latency and cost per successful task
• Familiarity with multiple model providers (Anthropic, OpenAI, Google, open-weight) and their trade-offs
• Experience with permissions, sandboxing and approvals in enterprise settings
• Nice to have: simulated-user evals, synthetic test generation, inference routing/caching, streaming or ML/RL/multi-agent research exposure

Apply now or send your CV directly!
They're moving fast, with only two places available. If you're a top-tier engineer who wants real ownership at the start of something big, get in touch today to start the conversation.