# Delta AI Documentation

> Delta AI is an architecture-first backend generator. Describe your product and it designs a backend system, writes the specification for every service, and opens the implementation as pull requests in your own GitHub organisation.

Delta AI is for backend engineers and technical founders. It generates TypeScript, Python, Go and Rust services backed by PostgreSQL, delivered as pull requests in your own GitHub organisation. You deploy the containers. Managed environments are not available yet.

## Pages

- [Welcome](https://docs.delta-ai.dev): What Delta AI does, what it produces, and how to read this documentation.
- [Onboarding](https://docs.delta-ai.dev/onboarding): The ten-step tutorial that appears in a new workspace, and what each step teaches.
- [GitHub App installation](https://docs.delta-ai.dev/github-app): Why Delta AI needs a GitHub App, what permissions it requires, and how to fix an installation that has drifted.
- [Create your first project](https://docs.delta-ai.dev/create-your-first-project): From a one-paragraph product description to an approved architecture, step by step.
- [How to build a backend](https://docs.delta-ai.dev/how-to-build-a-backend): The full workflow from an approved architecture to merged pull requests, and what happens at each step.
- [Plan mode](https://docs.delta-ai.dev/plan-mode): Pause any workflow step for approval, and steer a plan with a message instead of restarting.
- [Provider research](https://docs.delta-ai.dev/provider-research): How Delta AI researches external services it does not build itself, and how you choose between them.
- [Basics of modern backends](https://docs.delta-ai.dev/basics-of-modern-backends): The architectural vocabulary Delta AI uses, and why the generated systems look the way they do.
- [Sessions and chat modes](https://docs.delta-ai.dev/sessions-and-chat-modes): How sessions, conversations and chat modes relate, what reset does, and how parallel sessions work.
- [Change sets and merging](https://docs.delta-ai.dev/change-sets): How Delta AI groups pull requests into a change set, and how merging works.
- [Review and approve](https://docs.delta-ai.dev/review-and-approve): How to read the diffs Delta AI produces, and what approving a pull request actually does.
- [Credits and billing](https://docs.delta-ai.dev/credits-and-billing): How credits are charged, how holds work, and what happens when a run fails or is queued.
- [Pricing](https://docs.delta-ai.dev/pricing): Plans, credit costs, and what is and is not included.
- [Use case](https://docs.delta-ai.dev/use-case): A complete backend system built with Delta AI, published openly so you can read the output.
- [Glossary](https://docs.delta-ai.dev/glossary): The terms Delta AI uses, defined in one place.
- [FAQ](https://docs.delta-ai.dev/faq): Short answers to the questions that come up most often.
- [Troubleshooting](https://docs.delta-ai.dev/troubleshooting): What to do when a run fails, a merge is blocked, or the workflow does not behave as expected.
- [Support](https://docs.delta-ai.dev/support): How to reach us, and what to include so we can help quickly.
- [Changelog](https://docs.delta-ai.dev/changelog): What changed in Delta AI, newest first.

## Notes

- Pricing: the first architecture and one full service are free. Credits from $10, never expire, and cost less per credit inside a plan. Plans: Discovery $25, Pro $50, Business $200 per month with matching credits.
- Hosting and deployment are not included. Managed environments are coming.
- Every page is available as Markdown by appending `.md` to its URL.
- Full documentation text: https://docs.delta-ai.dev/llms-full.txt