Get a project
Create or join a Keboola project, find your way around it, and confirm Kai is switched on before you load any data.
Everything you build lives inside a project: the data you load, the SQL you write, the flow that runs them. This page gets you one.
Where you are: nothing exists yet. When you finish: a project you can sign in to, and you know where its Storage, Jobs and Components live.
Before you start
- Nothing else. No installs, no credentials, no files to download.
- An email address, for the project itself.
If your company already uses Keboola, ask a colleague to invite you instead of creating your own project; you probably want to work in theirs.
No prompt here. Kai works inside a project, so this one page is yours. Every page after it offers three ways to do the same task.
Get a project
Section titled “Get a project”Three ways in, in order of how quickly they get you started:
- Free Plan: sign up and create a free project right now. It is enough for everything in this guide: 250 GB of Storage and up to 60 minutes of job runtime a month (the free allowance tracks your previous month’s usage). See Pay As You Go for the full list of limits, including that the Free Plan does not let you choose a stack and gives you one project.
- Join an existing project: ask an admin of your company’s Keboola organization to invite you. They do it from Users & Settings in the project, with Invite User.
- Talk to us: for a trial of a full project with a chosen stack and no Free Plan limits, contact Keboola.
A project is a self-contained workspace with its own data storage, configurations, users and billing. Nothing leaks between projects unless you deliberately share it through the Data Catalog. Each project lives on one stack, a full instance of Keboola in one cloud region; you do not pick one here, because self-serve projects are created on a fixed stack. The list is in the platform overview.
Find your way around
Section titled “Find your way around”After signing in you land on the project dashboard. Five places matter for this guide:
- Components holds every data source connector, data destination connector, and application you can add. This is where you configure how data gets in and out.
- Transformations is the code that reshapes your data: SQL, Python, R, and dbt. Which of them a project offers is listed in its own Create Transformation dialog, so that dialog is the answer to “what can this project run?”.
- Storage lists every table and file, organized into buckets. Its tabs cover Tables & Buckets, Files, Data Streams, Storage Jobs and Events.
- Flows is the automation layer: what runs, in what order, on what schedule. Projects that have both kinds show this as Conditional Flows.
- Jobs is the log of everything that has run, successfully or not.
Check it worked
Section titled “Check it worked”Open Storage and you can see the project’s bucket list. Empty is fine; that is what you are about to change. And Kai Agent opens a chat panel when you click it. If the button is missing, or clicking it asks you to switch the feature on, that is the feature gate: an organization admin does it, and a non-admin needs them or Support. What Kai does also counts against the project, and an organization admin can cap how much it spends; Kai pricing and limits holds the current rules.
Not having Kai does not block the build: the loading, the SQL and the flow each carry a UI path and a CLI path in full.
If it goes wrong
Section titled “If it goes wrong”- The sign-up email never arrives. Check spam, then try again. A half-finished registration does not block a retry with the same address.
- You sign in and see no project. You are on the wrong stack, or your invite is still pending. Confirm the exact URL with whoever invited you.
- Your company uses SSO. Sign in through your organization’s link rather than creating a new account; a fresh Free Plan project will not have access to your team’s data.