Technical Product Lead (AI)

You were an engineer first. Then you started deciding what got built.We're hiring a Technical Product Lead, AI for a venture-backed LLM research startup. London, hybrid one or two days a week. £80,000 plus EMI equity options.Most product roles ask you to influence. This one hands you the keys, and the founders want them held by someone who has shipped systems, not just specs.THE COMPANYWe're Alteam, and we're helping a venture-backed LLM research startup hire its first Product Lead. It's a spinout from one of the world's top universities, founded by a leading AI researcher and a serial entrepreneur, so the science runs deep and things still ship fast.Their product replays your agent's real production traffic against candidate models and tells you exactly what changes before you switch, so enterprises can move onto cheaper models with evidence instead of hope. The buyers are engineers. The problem is technical. So are the product decisions.They're still finding product market fit and moving quickly. You'll spend three months building context with the founders making the final calls, then roadmap, scope and priorities become yours. There's a clear path to Head of Product or CPO.What you decide is what gets built, and you'll watch it land with real enterprise customers within weeks.THE ROLEOwn the roadmap and the outcomes, including the ones that don't workRun your own discovery with technical buyers and turn it into testable ideasKeep Product, Science and Engineering moving together, priorities clear and decisions madePrototype fast, then ship with EngineeringReport straight to the CEOYOUR PROFILEYou came up through engineering and moved into product because you wanted to decide what got built. You can hold your own with research scientists. In practice:You can open the trace of a failing multistep agent and form a hypothesis before you ask EngineeringYou design the eval set yourself, and you know where LLM as judge quietly falls apartYou've built multi-agent systems that real users depend on, with the guardrails that come with it: conclusions cited to source, scoped access, structured context between agentsYou can argue model routing, caching and distillation as cost levers, with numbers behind youYou build with agentic coding tools daily, Claude Code or Cursor or your own setup, and ship working agents from a spec rather than writing a doc for someone else to buildREQUIREMENTS (do not apply if you dont meet them)An engineering background. You shipped production code and owned real systems before moving into product8+ years in total, with genuine 0→1 product ownership since the switch, as a founding PM, early stage product lead, former founder or technical consultant who owned what got built2-3 years hands on with LLM agents in production: APIs, agentic systems, evaluation, traces and real failure modesHappy to take messy problems to a decision on evidence, without needing everyone to agreeEnterprise, regulated or otherwise high-stakes deployment is a strong plus, as is evaluation or observability workTHIS PROBABLY ISN'T FOR YOU IFYour AI work has mostly stayed in demos and pilots that never reached productionYou picked up the vocabulary without ever shipping the systemsYou've owned roadmaps but never a technical tradeoffYou want the authority without owning the outcomeSALARY AND PERKS£80,000 base salaryGenerous stock option package (EMI)5 weeks paid holidayFlexible schedule, hybrid one or two days a week in LondonImportant: the role is hybrid in London. You will need the right to work in the UK. No VISA sponsorship. Sound like you? Apply here or send it to info@alteam.io. Include why you fit, your Github handle and any example of work in Agentic AI. It's a highly competitive position.