← Back to list
Job

AI Product Engineer - ClickStack

AI / ML Engineer • Remote • Full-time • 📍 United Kingdom (remote)

AI Product Engineer role at ClickHouse building agentic capabilities on top of ClickStack, a petabyte-scale open-source observability platform (logs, metrics, traces, session replays). The focus is on agents that investigate incidents, propose root causes, and improve developer experience. You own the agent stack end-to-end in production, including context engineering, tool design, evals, tracing, and cost.

Responsibilities

  • Build agents that investigate incidents, surface anomalies, and answer 'why is production broken?' using ClickStack as their substrate
  • Write reusable skills (not just prompts) capturing how the team debugs, finds root causes, writes ClickHouse queries, and runs incident response
  • Own the agent stack end-to-end: context engineering, tool design, evals, tracing, and cost in production
  • Build MCP servers, SDKs, and integrations so customers' agents can read telemetry, take action, and stay observable
  • Work in the open with OSS contributors and customers, debugging their problems and feeding learnings back into the product
  • Tackle hard problems: latency, cost, context window limits, eval coverage, and hallucinations on real telemetry

Requirements

  • 5+ years of software engineering experience, including 1-2 years on LLM-powered systems or agents in production
  • Strong backend skills in TypeScript/Node.js and/or Python, comfortable in both
  • Hands-on experience building agents: multi-step tool use, planning, memory, error recovery
  • Experience designing skills (Markdown-based workflow encoding)
  • Production mindset: p99 latency, cost per task, long-term reliability

Nice to have

  • Experience wiring up MCP servers
  • Strong sense of good developer experience (DX) and developer tooling
  • Experience hitting and solving limits of generic copilots in production

Soft skills

Fast execution and frequent shippingComfort with ambiguity and ownershipStrong opinions on agent architecture from experienceLearning from failures
Languages: angol