Connecting data
How Flax connects to the data warehouse you already own, the warehouses it supports, and where you manage connections and credentials.
Flax is warehouse-native: it does not copy your data into a proprietary store. Instead you connect the warehouse you already own, and Flax compiles and runs queries directly against it. This page explains the connection model and points you at the per-warehouse setup guides.
#The connection model
A connection is a configured link to one data source. When you build a query in Explore, Flax compiles it to dialect-correct SQL and pushes it down to the connected warehouse for execution (ADR-0004) — your data stays where it lives, and you keep paying only your own warehouse for compute.
Every connection stores two things: non-secret configuration (host, database, and so on) and one secret (a password, token, or key). The secret is encrypted at rest with a rotating key-ring; see Security.
#Supported sources
| Source | Auth | Setup |
|---|---|---|
| Postgres | User + password | Postgres |
| Snowflake | Key-pair (RSA) | Snowflake |
| Databricks | Personal access token | Databricks |
| BigQuery | Service-account JSON | BigQuery |
| File upload | — | CSV uploads |
If you do not have a warehouse — or just want to explore a file quickly — you can upload a CSV and query it directly. See CSV uploads.
#Where you manage connections
Warehouse connections live on the Data page at /app/data. From there you can add a connection, run a live connection test, and edit or remove existing ones. Uploaded files are managed separately on the Uploads page at /app/uploads.