Setup guides
Get your data where Dataki can ask it things
Dataki connects to BigQuery, PostgreSQL, MySQL, Amazon Redshift, Supabase, Google Sheets and files directly. Everything else reaches it through one of those. Each guide sets up one route, step by step, checked against the provider's own documentation, and says where it stops.
01 Someone else's export
Through BigQuery
Google's products, Firestore and Stripe all have a way into BigQuery. Set it up once, connect the dataset, and Dataki queries it like any other warehouse.
-
Google Analytics 4 Link GA4 to BigQuery, connect the export to Dataki, and ask about users, sessions, pages and purchases in plain English. Ten minutes of setup; data from the next day. Google Analytics 4 BigQuery export Dataki -
Firebase Turn on Firebase's BigQuery export for Analytics, Crashlytics, Performance Monitoring and Cloud Messaging, connect the datasets to Dataki, and ask about users, releases and crashes in plain English. Firebase BigQuery export Dataki
-
Google Search Console Set up Search Console's bulk data export to BigQuery, connect the dataset to Dataki, and ask about queries, clicks, rankings and pages in plain English, with no 1,000-row cap. Search Console Bulk data export Dataki
-
-
Stripe Copy Stripe charges, customers, invoices and subscriptions into BigQuery with Google's Stripe connector, connect the dataset to Dataki, and ask about revenue in plain English. Stripe BigQuery Data Transfer Service Dataki
02 A connection you hold
Your own database
Any PostgreSQL or MySQL server, Supabase project, Amazon Redshift warehouse or BigQuery project Dataki can reach. What differs from host to host is where the connection details live and who is allowed in.
-
Hosted PostgreSQL Create a read-only Postgres user, allow Dataki's fixed IP, and connect any hosted PostgreSQL: Neon, Amazon RDS and Aurora, Cloud SQL, Azure, Render, Railway, Heroku, DigitalOcean and more. Your Postgres host Direct connection over TLS Dataki -
Supabase Sign in with Supabase, pick the project, and Dataki fills in its pooler connection. Add the password, or a read-only role that row level security lets read. Every step checked against Supabase's documentation. Your Supabase project Shared pooler over TLS Dataki -
Hosted MySQL Create a read-only MySQL user, allow Dataki's fixed IP, and connect any hosted MySQL: PlanetScale, Amazon RDS and Aurora, Cloud SQL, Azure, DigitalOcean, TiDB Cloud and MariaDB. Your MySQL host Direct connection over TLS Dataki -
Amazon Redshift Create a read-only Redshift user, open the cluster or Serverless workgroup to Dataki's fixed IP on port 5439, and connect it. Dataki writes Redshift SQL and reads every schema the user can see. Your Redshift cluster or workgroup Direct connection over TLS Dataki -
Google BigQuery Link a Google Cloud project to Dataki. It creates a service account there that can read tables and run queries, every dataset becomes a source, and every query runs, and is billed, in your project. Your Google Cloud project Dataki's service account Dataki
03 A copy you upload
A file
When there is no live route, an export uploaded as CSV or Excel is queried with full SQL.
Something else?
If a sync tool, BigQuery's Data Transfer Service or a CSV export can put it in a database or a warehouse, Dataki can read it. Google's transfer service alone covers Salesforce, HubSpot, Shopify, Google Ads and more. Tell us what you are trying to connect and we will point you at the shortest route.