Job Description:

AI Full Stack Engineer | LLM Products on AWS

Central London (In person) | Permanent

We're hiring on behalf of an AI-first analytics startup. Their product is built on a production multiagent system that uses Anthropic, OpenAI and Perplexity models, runs on AWS, and reaches customers through the company's own MCP server. W

e're looking for a strong full-stack or product engineer who has moved into building with LLMs and wants to own the whole AI system, end to end: the agents, the evals that prove they work, the data pipelines behind them and the cloud they run on.

What you'll do

• Build and ship LLM-powered features from idea to production, across agents, APIs and infrastructure
• Extend the multi-agent platform: tool calling, structured outputs, memory and context management
• Add new tools to the company's MCP server Build evals and tracing so the team can ship AI features quickly and safely Run and improve the AWS stack (ECS Fargate, Lambda, Step Functions, S3, DynamoDB, Bedrock), with CI/CD and infrastructure as code

What we're looking for

• 3–5 years as a software engineer, ideally full stack or backend, including real experience shipping
• LLM or GenAI features to production users Strong Python Hands-on with LLM APIs (tool calling, structured outputs) and at least one agent framework (LangGraph, CrewAI, OpenAI Agents SDK, Claude Agent SDK or similar)
• Comfortable on AWS and with event-driven, distributed systems A bonus: RAG, LLM evals, MCP, or AI side projects and open-source work you can show us
• Right to work in the UK

If you've built AI features end to end and want to own the system behind them, not just the screen in front, we'd like to hear from you.