LONTARI is an open initiative building a shared, living data model for coffee — from plot and lot to roast batch and shipment. The goal is a market that's easier to trust, for everyone who touches it.
Origin, quality, and movement data for coffee lots sit scattered across trader systems, cooperative ledgers, roasters, and regulators — each with its own format and its own version of the truth. LONTARI builds a shared, verifiable ontology: a common vocabulary of entities and relationships, where a plot, a lot, a roast batch, and a compliance declaration are nodes in one living graph, not rows in someone else's spreadsheet.
The schema, entity dictionary, and linking rules are developed in the open, available to anyone who wants to check them.
The model treats a claim and a measurement as distinct types of data. Every statement about a lot traces back to a source — a document, an instrument, or a person.
LONTARI doesn't sell participant data and doesn't take sides. Development is funded by grants, the community, and direct contribution — not by any single market player.
Two working hypotheses that shape what we build first — working assumptions that guide priorities, not a financial forecast.
The first version of LONTARI focuses on the core provenance graph: uploading and reconciling lot data, checking claims against measurements, and a read-only view of the ontology. The alpha is meant for first-pass analysis, not production use — we'll be upfront about its limits as we run into them.
The underlying graph model isn't just a whiteboard exercise — it's already tracking live lots inside working roastery deployments, on-premise and offline by design. The alpha brings that same model to the wider community.
LONTARI moves forward on contributed time and expertise. A few ways to help right now.
Data engineers, backend developers, and ontologists — we need hands on the graph model and reconciliation tooling.
Producers, traders, and roasters willing to share lot data on their own terms.
Agronomists, compliance and logistics specialists to stress-test the model against real cases.
As a non-profit initiative, LONTARI welcomes grant and partner funding for development.
