Why I'm building Airdash.

The problem

If you work with AI agents like Claude or Codex, you ask them to keep track of things: leads, sources, tasks, expenses. Then the list ends up in a chat history, or in a markdown file that the agent rewrites in full every time.

That works for ten items. It stops working at a few hundred, and you cannot easily look at it, sort it or fix one value.

What Airdash is

Airdash gives the agent a real database. Through an MCP server, the agent creates a Cloudflare SQLite (D1) database, adds tables and columns, and fills them. For you, each database has a clean, Airtable-like view, so you can read the data and edit it too.

Agents are the main users. Anything you can do in the UI, an agent can do through MCP.

Why you might use it

• You use AI agents daily and want them to keep structured notes.
• You want to see and correct what the agent stored, without learning SQL.
• You use more than one assistant and want one place for the data.
• You prefer a small, plain tool with published prices: from $79 per month.

Why you might not

Airdash is new and in early use by me. It is not a full Airtable replacement with forms, automations and dashboards. If you need those, use a tool built for them. If you only want a database your agent can use and a grid you can look at, that is what Airdash is for.

Who is behind it

Airdash is built by me, Peter Van Dijck, at Helpful Intelligence, in Madrid, Spain. If you have a question or a feature you need, write to hello@getairdash.com. It goes to me.