Catalog
Product Information Management on the EAV substrate
What it does
A PIM where every product attribute is a first-class EAV cell — sparse, typed, and directly traversable by Fundi agents. Senses connect upstream to enrichment providers and to GS1 for standards-aligned identification, so the catalogue stays current without manual data wrangling. The same graph that holds product attributes carries memory for the agents reasoning over it, so search, classification, and recommendation share one substrate.
Where most PIMs sit beside the database that runs the rest of the business, Catalog is a view of the typed-memory substrate itself. A product attribute is the same primitive as any other typed object, so the same Fundi that classifies a catalogue item can also walk its supplier relationships, its certifications, its events, and its workflows.
Key features
- PIM on the EAV substrate — every attribute first-class
- Senses integrate enrichment providers and GS1
- Native agent traversal, no separate search index
How it works
Products live as typed objects on the Object Store; attributes are EAV cells on those objects. Senses wire upstream to enrichment providers and to GS1 — the global identifier standards body — so attributes flow into the catalogue as traversable nodes instead of imported rows. Agents read product attributes via the same MCP contract they use for any other memory walk.
- Object Store — the EAV substrate every catalogue attribute lives in.
- Senses architecture — how enrichment providers and GS1 are wired in.
- Memory — the access layer agents use to reason over product attributes.
Try it
Ingest your first catalogue feed via the quickstart.