AI Engineer

💻 Job Title: AI Engineer💰 Salary: up to ~£90k (potential flex) + early stage equity📍 Location: could be fully remote (Central London office if preferred)🗣️ Industry: Conversational AI / RecTech🛂 VISA sponsorship is NOT availableA young, founder-led AI company is hiring an AI Engineer to join a small, high-calibre team building conversational AI for the recruitment industry. Not a research role. Not a pilot. Live AI products, used by real customers and you’ll be working on the part of the product where the AI logic actually lives.You’ll work directly with the AI lead, alongside full-stack engineers who own the integrations and APIs. That gives you the space to focus on what makes the agents good: quality, testing and experimenting. Engineers here get a lot of freedom and you’ll be involved in product decisions and close to customers.🛠 What you’ll work on:A conversational agent that talks to candidates throughout their journey: engagement, collecting documents, scheduling and staying in touch after placement.A copilot that lets recruiters search their candidate data in plain language, calling tools and surfacing insights along the way. You’ll help build the next version.In-product agents that guide users through setting up their own workflows.Evaluating and improving agent quality, with prompts that are written by hand, not generated.You’ll have real ownership of what you build, from first idea to production.✅ What we’re looking for:3+ years in software or AI engineering, with strong PythonExperience shipping LLM-powered products built on third-party models (OpenAI, Anthropic, Gemini). No model training needed.Hands-on with agents and tool calling, with a solid grasp of how LLMs and agents use toolsNLP and search fundamentals: embeddings, semantic search and retrievalStrong ownership: a self-starter who takes initiative and drives work to the end without being managedA product mindset: you understand how your work fits the wider product and what matters to customersComfortable in an engineering team: GitHub, PR review, testing and knowing when something is ready to shipAble to write and iterate good prompts by hand and judge whether the output is actually good👍 Nice to have:Agents built and deployed end to endVector database experience (Weaviate, Pinecone, pgvector or similar)Conversational AI experience (chat or voice agents), such as multi-turn context, dialogue design or real-time interactionSome full-stack work (React/TypeScript) and comfort stepping outside the AI layerAgent frameworks, FastAPI and LLM observability tooling