# Union.ai

Union.ai empowers AI development teams to rapidly ship high-quality code to production by offering optimized performance, unparalleled resource efficiency, and a delightful workflow authoring experience. With Union.ai your team can:

* Run complex AI workloads with performance, scale, and efficiency.
* Achieve millisecond-level execution times with reusable containers.
* Scale out to multiple regions, clusters, and clouds as needed for resource availability, scale, or compliance.

> [!NOTE]
> Union.ai is built on top of the leading open-source workflow orchestrator, [Flyte](https://www.union.ai/docs/v1/flyte/).
>
> Union.ai provides [all the features of Flyte, plus much more](https://www.union.ai/docs/v1/union/user-guide/introduction/page.md),
> available as a fully managed BYOC (Bring Your Own Cloud) service or as a self-managed deployment on your own infrastructure.
>
> You can switch to another product version with the selector above.

### [Introduction](https://www.union.ai/docs/v1/union/user-guide/introduction/page.md)

Union.ai builds on the leading open-source workflow orchestrator, Flyte, to provide a powerful, scalable, and flexible platform for AI applications.

### [Getting started](https://www.union.ai/docs/v1/union/user-guide/getting-started/page.md)

Build your first Union.ai workflow, exploring the major features of the platform along the way.

### [Core concepts](https://www.union.ai/docs/v1/union/user-guide/core-concepts/page.md)

Understand the core concepts of the Union.ai platform.

### [Development cycle](https://www.union.ai/docs/v1/union/user-guide/development-cycle/page.md)

Explore the Union.ai development cycle from experimentation to production.

### [Data input/output](https://www.union.ai/docs/v1/union/user-guide/data-input-output/page.md)

Manage the input and output of data in your Union.ai workflow.

### [Programming](https://www.union.ai/docs/v1/union/user-guide/programming/page.md)

Learn about Union.ai-specific programming constructs.

### [Administration](https://www.union.ai/docs/v1/union/user-guide/administration/page.md)

Union.ai administrators can manage users, projects, and resources.

### [Integrations](https://www.union.ai/docs/v1/union/user-guide/integrations)

Union.ai integrates with your cloud resources and external services.

### [FAQ](https://www.union.ai/docs/v1/union/user-guide/faq/page.md)

Frequently asked questions.

## Subpages

- [Introduction](https://www.union.ai/docs/v1/union/user-guide/introduction/page.md)
  - Flyte
  - Union.ai
  - BYOC (Bring Your Own Cloud)
  - Self-managed
- [Getting started](https://www.union.ai/docs/v1/union/user-guide/getting-started/page.md)
  - Gather your credentials
  - Log into Union.ai
- [Core concepts](https://www.union.ai/docs/v1/union/user-guide/core-concepts/page.md)
  - Defining tasks and workflows
  - Type annotation is required
  - Workflows *are not* full Python functions
  - Registering tasks and workflows
  - Registering on the command line with `union` or `uctl`
  - Registering in Python with `UnionRemote`
  - Results of registration
  - Changing tasks and workflows
  - Inspecting tasks and workflows
  - Inspecting workflows in the UI
  - Inspecting tasks in the UI
  - Inspecting workflows on the command line with `uctl`
  - Inspecting tasks on the command line with `uctl`
  - Inspecting tasks and workflows in Python with `UnionRemote`
  - Running tasks and workflows
  - Running a task or workflow in the UI
  - Running a task or workflow locally on the command line with `union` or `python`
  - Running a task or workflow remotely on the command line with `union`
  - Running a task or workflow remotely in Python with `UnionRemote`
- [Development cycle](https://www.union.ai/docs/v1/union/user-guide/development-cycle/page.md)
- [Data input/output](https://www.union.ai/docs/v1/union/user-guide/data-input-output/page.md)
- [Administration](https://www.union.ai/docs/v1/union/user-guide/administration/page.md)
- [Programming](https://www.union.ai/docs/v1/union/user-guide/programming/page.md)
- [FAQ](https://www.union.ai/docs/v1/union/user-guide/faq/page.md)
  - Onboarding my organization to Union.ai
  - What information does Union.ai need to set up my service?
  - How do I change the machine types in my cluster?
  - Data storage and handling
  - How does Union.ai store my data?
  - Can I change the raw data storage location?
  - Can I use my own blob store for data storage that I handle myself?
  - Can I control access to my own blob store?
  - Could someone maliciously delete or otherwise access my raw data?
  - Can I use s3fs from within a task?
  - Can I use BigQuery from within a task?
  - `FlyteFile` and `FlyteDirectory`
  - Where do `FlyteFile` and `FlyteDirectory` store their data?
  - Can I accidentally overwrite `FlyteFil`e data?
  - Can I use my own blob store for `FlyteFile` and `FlyteDirectory` data storage?
  - How do the typed aliases of `FlyteFile` and `FlyteDirectory` work?
  - Building and running workflows
  - What SDK should I download and use in workflow code?
  - How do I authenticate `uctl` and `union` CLIs to Union.ai?
  - How do I specify resource requirements for a task?
  - What command-line tools should I use to register and run workflows?
  - How do I fix import errors when running workflows remotely?
  - What happens if an automated process launches a very large number of workflows?
  - How can I constrain the number of parallel executions for large, complex workflows?

---
**Source**: https://github.com/unionai/unionai-docs/blob/main/content/user-guide/_index.md
**HTML**: https://www.union.ai/docs/v1/union/user-guide/