> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.umnai.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.umnai.com/_mcp/server.

# Get started

The platform is an API-first solution for building, deploying, and querying explainable models.

Use the platform to move from source data to deployed Hybrid Intelligence models that return both predictions and the explanation data behind them.

## How to access the platform

Organisations onboarding and account creation are handled by Umnai. Once your account is available, you can start using the platform.

## How the platform works

Most workflows move through the same path:

Grant the platform access to source data in your storage environment.

Convert source files into a validated **datasource** managed by the platform.

Transform the datasource into a training-ready **dataset**.

Train a **model** from the dataset.

Create a **deployed model** that can receive API requests.

Query the deployed model through a **view**, which returns predictions, explanations, diagnostics, or evaluation outputs depending on the selected view type.

## Start building

The fastest way to understand the platform is to run one complete workflow with sample data.

Ingest sample data, train a model, deploy it, and request your first prediction and explanation.

Look up endpoints, request fields, response schemas, and examples.

Implement each stage of the platform lifecycle in more detail.

Learn the model, explanation, and Hybrid Intelligence concepts behind the platform.