Common Data Source & Logic Patterns

Introduction

PromptQL, powered by Hasura DDN, makes it simple to build AI-ready APIs on your data — no matter where it lives.

You'll connect your data using native connectors, configure your API using Hasura Metadata Language (HML), and create an immutable build that will serve your application.

Then, you'll use the PromptQL Playground UI or the PromptQL API to chat with your data, ask questions, or automate tasks — just like you would with a trusted analyst or engineer who speaks your business’ language.

Let's build something

To get started, it's helpful to look at common patterns for building PromptQL applications based on the datasource(s) you want to include:

• With a persistent datasource and custom business logic
• With API endpoints
• With bulk API data

Mix and match

Keep in mind: these are simply broad buckets for you to get started. Any sources we support can be added to the same project to create a unified agentic semantic metadata layer which feeds data to PromptQL.

Guides by data source

You can also dive straight into a guide for your preferred source:

• Amazon Athena
• Amazon Redshift
• BigQuery
• Databricks
• MySQL
• PostgreSQL
• Snowflake