How it works
Application model and pipeline
This page maps Klaaro's core entities and the processing pipeline that runs on every upload. Use the application model to see how teams, datasets, documents, and records relate in the dashboard and /api/v1 — then follow the Document Engine diagram for the ordered steps from ingest through extraction.
Application model
A team is the top-level tenant: members, roles, and API keys. Each dataset belongs to one team and holds pipeline configuration (classes and per-type schemas), ingested documents, and the records extracted from them. Classes label document types; schemas define the fields extraction writes into records.
Document Engine
Upload triggers the pipeline without extra API calls. Each document moves through parse and OCR, then classification and extraction against the dataset's classes and schemas. Results surface in View (grid, insights, review) and as structured records via the API. Re-run from any step is possible.
Runs end-to-end after every upload — re-run from any step