AI Product Engineer (Backend, Node.js)

Smeetz is an AI-powered unified commerce platform for the European leisure and entertainment industry. Theme parks, family entertainment centres, museums, zoos, and theatres run their ticketing, seating, F&B, and payments on us.
This role owns the backend of our two AI surfaces: the Smeetz MCP server and the AI chatbot. Both are backend services at heart, made of APIs, integrations, and agent logic sitting on top of our commerce platform. Expect to spend about 70% of your time on backend engineering (building, scaling, and running these services in production) and about 30% on the product side (understanding how venues use them, shaping what gets built next, and measuring whether it worked).
You report to the VP of Engineering and have no direct reports.
Key Responsibilities
Backend
Own the backend of the Smeetz MCP server and AI chatbot end to end, covering the services, the tool surface, the agent logic, and the guardrails on what they are allowed to do
Design and build the APIs and integrations that connect our AI surfaces to the Smeetz platform (ticketing, bookings, payments) with the security, performance, and reliability that production commerce requires
Own reliability and performance, covering latency, error rates, cost per call, and the call patterns clients actually generate. You get full observability data (Langfuse) and the mandate to act on it
Build the eval layer so prompt and model changes are measured before they ship, not after a customer notices
Participate in on-call for the services you own
Product
Work with the product owner and venue-facing teams to understand how operators use the AI surfaces, and turn that into technical specs and priorities
Read traces and usage data to spot what is breaking or missing, and propose what to build next
Define success metrics for what you ship and report back on them
Document the surfaces you own so the next engineer can work on them without you in the room, and extend our shared Claude Code skill library
Requirements
Professional requirements
5+ years shipping production code, including 3+ years of backend engineering building and running production APIs. Node.js/TypeScript is strongly preferred
Solid backend fundamentals, covering API design, data modelling, authentication and permissions, queues and async processing, and debugging under production load
You have shipped at least one LLM-based feature to real users and worked on its backend side, including tool calling, structured outputs, cost and latency control, and production failure modes. Prototypes and demos do not count
Ownership of a production system, including the on-call that comes with it
Daily use of Claude Code or an equivalent agentic environment beyond autocomplete
Comfortable talking to non-technical stakeholders and turning their problems into technical work
Fluent English
Nice to have
Designing a tool or API surface for agents, with naming, granularity, and error messages a model can recover from
The MCP specification, beyond using it as a client
The Anthropic Claude API and tool use
Langfuse or a similar tool (LangSmith, Braintrust, Phoenix)
Evals and retrieval over a support knowledge base (Intercom Fin experience is relevant)
Guardrails for agents with write access, covering permissions, confirmation steps, prompt injection, and PII
Experience in a product-engineering setup where you shaped priorities, not just tickets
French or German
Personal attributes
Owner: The services you run work because you are responsible for them
Backend-first, product-aware: Starts from solid engineering, and knows why the feature matters to the venue
AI-native: Uses Claude and agents as everyday leverage
Measured: Defines success before building and accepts the number
Writer: Leaves the codebase better documented than they found it
Benefits
A published dual career ladder. Grow to Staff AI Product Engineer without becoming a manager, on pay bands comparable to the management track
Full ownership of two AI surfaces that are core product, not a side project
Direct contact with the venue operators who use what you build
Modern AI-native tooling: Claude Code with a shared internal skill library
Fully remote position
Competitive salary based on experience. Band to confirm

Work arrangement
Yes