Dataset Onboarding
This first slice imports a CSV, profiles it deterministically, drafts a semantic contract, and stores the normalized rows in Postgres before you start querying.
Step 1
Upload
Step 2
Inspect
Step 3
Semantic Draft
Step 4
Complete
Current step
Limits: 25 MB, 100k rows, 200 columns, UTF-8 CSV with a header row.
What this slice does
Imports a CSV and stores normalized rows in Postgres.
Profiles types, identifiers, measures, dimensions, and the primary time field heuristically.
Generates a semantic draft that QueryLens uses as the safe contract for first queries.
Limits supported questions to discovery, aggregates, grouped summaries, and time trends.