# Turn heart signals into on-device intelligence

## What is heartKIT?

heartKIT brings signal preparation, model development, evaluation and export into one Python workflow for ECG and PPG applications. Start with a built-in task and configuration, then change the data or model to fit the question you are working on.

You can run a configured workflow from the command line or use the Python API to assemble one yourself. The task pages explain each application; the guides walk through examples, and the API reference covers the underlying interfaces.

## Explore heart tasks

Choose a starting point for the signal or question you want to work with.

Denoising
Clean ECG and PPG recordings before analysis.
Explore denoising →

Segmentation
Locate regions and events in a waveform.
Explore segmentation →

Rhythm
Build an ECG rhythm-classification workflow.
Explore rhythm →

Beat classification
Classify individual heartbeats in an ECG recording.
Explore beat classification →

## Installation

In a Python 3.12 project, add heartKIT and check that the command line tool is available:

```bash
uv init --python 3.12 my-heart-project
cd my-heart-project
uv add heartkit
uv run heartkit --help
```

The [quickstart](https://ambiqai.github.io/heartkit/quickstart/) covers other installation methods and your first workflow.

## Explore the documentation

- [Getting started](https://ambiqai.github.io/heartkit/quickstart/)Install heartKIT and run a first example.Start here →

- [Tasks](https://ambiqai.github.io/heartkit/tasks/)Choose an application and follow its workflow.Explore tasks →

- [Guides](https://ambiqai.github.io/heartkit/guides/)Work through practical examples and notebooks.Follow a guide →

- [Python API](https://ambiqai.github.io/heartkit/reference/)Look up packages, classes and function parameters.Browse the API →