# Connect your first data source
Open **Connectors**, pick the tool, choose a project and send the request. Your Studio's AI drafts the connection; you finish it with your own key. Then **Test** it and load data into a table.
## What you will build [#what-you-will-build]
A tested connection from a project to a tool such as Google Sheets, PostgreSQL or Amazon S3, and a table in Nexus filled from it.
GitHub is the one connector you connect directly, from Settings. See [GitHub](/docs/connectors/github).
## Before you start [#before-you-start]
- A project with a Studio you can open. See [Your first project](/docs/guides/first-project).
- An account on the tool you want to connect, and a key or sign-in for it.
- The project's **Internet access** must not be off, or connections do not work.
Keys you give a connection are encrypted. Not even the AI can read them.
## Steps [#steps]
### Find the connector [#find-the-connector]
Go to and search for your tool. Open it to see what it does and what you need.
### Ask your Studio's AI to connect it [#ask-your-studios-ai-to-connect-it]
Under **Connect it with your Studio's AI**, pick the project and send the request. Or, inside a project, open its connections, describe what you want (for example "my Stripe account") and select **Ask AI to connect**.
If the Studio is closed, the AI reads the request when it opens.
### Finish it with your key in the Studio [#finish-it-with-your-key-in-the-studio]
The AI drafts the connection and tells you where to finish it. Paste your own key or sign in where it asks.
### Test the connection [#test-the-connection]
In the project's connections, select **Test**. A working connection shows when it was last used. If it fails, select **Reconnect** and re-enter the key.
### Load data into a table [#load-data-into-a-table]
Ask the AI in the Studio to load what you need, for example:
> Load the last 90 days of invoices from Stripe into a table called invoices, and keep it up to date every night.
Then open to see the new table.
## Check that it worked [#check-that-it-worked]
- The connection appears in the project's connections, with the date it was last used.
- **Test** succeeds.
- The new table appears in **Nexus Lakehouse** with rows.
## Common variations [#common-variations]
- **Receive events with a webhook.** For services that push events: in the project's connections select **Add webhook**, name it, and give the service the address Klarety shows. Paste the service's signing secret so Klarety can check each event is real.
- **Share a connection.** A connection can belong to the workspace instead of one project. Give a teammate access in [Data access](/docs/data/data-access).
- **Pause a connection.** **Pause** stops using it; **Resume** starts again. **Remove** deletes the key, and every Studio using it stops reaching the service.
## If something goes wrong [#if-something-goes-wrong]
Connections need the Studio to reach the service. Change **Internet access** in the project's settings.
The Studio reaches only its connections and the sites you add. A new connection's service must be among them.
Only workspace admins connect GitHub. Ask an owner or admin to connect it under .
The key may have expired or been revoked at the service. Select **Reconnect** and enter a new key.
## Next steps [#next-steps]
- [Connectors](/docs/connectors): every tool you can connect.
- [PostgreSQL](/docs/connectors/postgresql), [Amazon S3](/docs/connectors/amazon-s3) and [Google Sheets](/docs/connectors/google-sheets): step-by-step setup.
- [Build your first agent](/docs/guides/first-agent): act on the data you brought in.