HomeInstallHow it Works
DocsBlogPricingPlatformInstall
HomeDocsQuickstart

Quickstart

Install Baseil, connect your first database, and run your first natural-language query in under 5 minutes.

4 min read

This is the fastest path from zero to a working Baseil instance with one database connected and chat responding to your questions. Budget around five minutes if you already have a database handy, a bit longer if you need to spin one up.

Prerequisites

You'll want these in place before you start:

  • ‣Python 3.12 or newer
  • ‣Node.js 18+ (only needed if you plan to build the web UI from source)
  • ‣Docker, optional, makes spinning up a local Postgres trivial. You can skip it if you already have a database to point Baseil at
  • ‣An Anthropic API key for Claude, which is the default LLM

Auth for local development runs through Clerk, and the setup wizard provisions everything you need. You don't have to create a Clerk account first.

Install the CLI

The simplest install is the published shell script:

curl -sSL https://releases.baseil.ai/install.sh | bash

This drops the baseil binary into your path and nothing else. No database changes, no system services. Uninstalling is just removing the binary.

Run setup

baseil setup

The setup wizard walks through six prompts. Each has a sensible default if you just hit enter:

  1. ‣Postgres detection. Baseil needs a Postgres instance for its own metadata (connections, rules, golden cache, audit logs). It looks for a local one first, then offers to start one via Docker if you have it. BYO connection string also works.
  2. ‣Database schema creation. Baseil creates its schema in the database you pointed it at. Tables are namespaced so this is safe to run on a shared database.
  3. ‣Admin account. Email and password for the first admin user. This is separate from the Clerk session and is what you'll use to sign into the web UI.
  4. ‣Clerk keys. For local dev the wizard can auto-provision a Clerk development instance. If you already have Clerk credentials, paste them in. See Clerk's docs if you want to manage this manually.
  5. ‣LLM provider. Anthropic (default) or OpenAI. Paste your API key here. This key never leaves your machine in local dev.
  6. ‣Embedding model. Used for the golden cache. Default is text-embedding-3-small on OpenAI, with a local fallback available.

A successful run looks like this:

> baseil setup

[1/6] Detected Postgres at localhost:5432 - OK
[2/6] Creating baseil schema... done
[3/6] Admin account: [email protected] created
[4/6] Clerk dev instance provisioned
[5/6] Anthropic key validated
[6/6] Embeddings: text-embedding-3-small

Setup complete. Run `baseil start` to launch the server.

Start the server

baseil start

This launches the backend API on port 8451 and serves the web UI from the same process. When it's up, you'll see:

Baseil is running at http://localhost:8451

Open that URL in a browser and sign in with the admin account you just created.

Your first query

You're one connection away from asking questions.

  1. ‣Click Connections in the sidebar.
  2. ‣Hit Add Connection and walk through the form. Any Postgres, MySQL, SQLite, or Elasticsearch will do. (If you need more detail, see Connecting Databases.)
  3. ‣Wait for onboarding to finish. Usually 30-60 seconds. The progress panel shows each stage of the pipeline.
  4. ‣Open Chat and ask something.

Good first questions to try:

  • ‣"How many rows are in the users table?"
  • ‣"Show me the most recent 10 orders"
  • ‣"What's the schema of the products table?"

Each response comes back with the natural-language answer up top, the tool that ran, the actual SQL, and a small metadata footer with row count and execution time. If the number looks wrong, you can verify it right there against the query.

What happened under the hood

When you added a connection, Baseil ran a five-agent pipeline: Discovery mapped your schema, ToolBuilder generated parameterized query templates, Reviewer checked them for injection risks and read-only enforcement, Tester ran them against your real data with safe parameters, and Deploy registered everything as tools the chat can use. The whole thing takes less than a minute for a typical schema.

For a deeper look, read Inside Baseil's 5-Agent Pipeline.

Next