Before it operates, the platform understands: it writes down how the thing it operates works. You build that map, the ontology, and the connectors that feed it from the client's systems.
The role
- Model assets, orders, routes, units, people and how they connect.
- Write connectors that read from the client's systems without taking their data off their side.
- Look after data quality: duplicates, gaps, contradictions between systems.
- Make agents query the ontology instead of guessing.
Your first 90 days
- Day 30
You learn the current ontology and its tests; you close a gap with its case.
- Day 60
You build a complete connector for a client system, with tests.
- Day 90
A client's ontology stays current on its own and the agents use it to decide.
Stack and tools
PostgresSupabaseTypeScriptPythonpgvector
And what our clients already use
What we look for
Must have
- Fluent SQL and a taste for data modeling.
- You have integrated systems that were not built to talk to each other.
- You notice when a number doesn't add up and you don't let it slide.
- You can explain a data model to someone in operations.
Nice to have
- Postgres, graphs or semantic modeling.
- Experience with the systems a large company runs on.
- Vector search or information retrieval.
What working here is like
What you build goes into products that operate real organizations. You work directly with the founders, with no layers between your code and production, under the same rule we give the intelligence: a person signs what is critical.