Founding AI Engineer

ABOUT THE ROLE

Bardin is the always-on application engineer for the robotics and automation industry. We're a Physical AI company: the engineering judgment required to scope, deploy, and support complex automation, captured in machine-readable form for the first time.

We've spent a year proving this works. The knowledge graph, agents, and data pipeline are live against real customer data. We know what breaks and what our users need. Now we're building out the enterprise-ready version.

You'll join as our Founding AI Engineer to take a working system and make it excellent: deeper ontology, more reliable agents, retrieval that holds under production load, evals that let us move fast without breaking trust. You inherit a year of hard-won context and own entire domains of it outright.

WHO YOU'LL WORK WITH

You'll work directly with the CTO, who's been in the graph and model layer since day one. You'll split systems work with our senior backend engineer, so the AI layer isn't slowed by infrastructure you carry alone. And you'll work daily with our industrial application engineer, who comes from the automation world.

Small team. Short feedback loops. No layers between you and the decision.

IN THIS ROLE YOU'LL

• Own and expand the industrial knowledge graph and ontology our agents reason over: entities, relationships, schema, and the ingestion that populates it from technical PDFs, spreadsheets, and scraped web content.
• Build stateful, multi-step agents that plan, call tools, and hold context, and architect the graph-based retrieval that feeds them.
• Stand up observability and evals (tracing, offline and online evaluation, feedback loops) so we improve agent reliability with evidence instead of intuition.
• Build the production backend it runs on: resilient APIs, event-driven and async patterns, real-time streaming, reliable job processing.
• Work directly with customers to learn how application engineers and integrators actually operate, and loop that back into the agents and the graph.

Requirements

• 6+ years backend engineering, fluent in TypeScript and Python, production systems with real users.
• 2+ years shipping LLM-powered products: retrieval, prompting, tool use, structured output, multi-step agent workflows.
• Strong ML fundamentals. NLP or applied ML background, comfortable with evals and non-deterministic systems.
• Hands-on with knowledge graphs or ontologies (Cypher, FalkorDB, Neo4j, RDF).
• Experience building agents and orchestration (Mastra, LangGraph, or similar).
• Solid data modeling over messy, unstructured inputs.
• Experience with Gemini, Vertex AI, or comparable platforms.
• Energized by inheriting a real codebase with real users, and by deciding what to keep, rebuild, or throw out.