EvalML
evalml.alteryx.com
- Free plan
- Windows
- macOS
- Linux
- API
- Self-hosted
What EvalML is
EvalML is a free AutoML library for building, optimizing, and evaluating machine-learning pipelines. It assembles pipelines with preprocessing, feature engineering, feature selection, and multiple modeling techniques, while automation features include data-quality checks and cross-validation. Users can select objectives such as mean squared error, cross entropy, or area under the ROC curve, and can define custom objectives for domain-specific work. Tools are included for understanding and inspecting models. EvalML can combine with Featuretools and Compose to form end-to-end supervised machine-learning solutions. Tutorials cover fraud prediction, lead scoring, cost-benefit objectives, and text data. Installation is available through PyPI, conda-forge, or source, with Python 3.9–3.11 listed on the current installation page. Optional XGBoost, CatBoost, Plotly, and ipywidgets dependencies extend modeling and plotting. Time-series support uses Prophet and remains under active development. The library supports Linux, macOS, Windows, API, and self-hosted use.
Who it suits
EvalML suits developers and data scientists who want code-based AutoML, custom objectives, model inspection, and automated pipeline selection. It is also relevant for supervised workflows using Featuretools or Compose.
Good to know, and what to watch
- Automated pipeline construction and optimization
- Feature engineering and model selection
- Custom and standard objectives
- Data-quality checks and cross-validation
- Model understanding tools
- Python support listed only for versions 3.9–3.11
- Mac requires OpenMP for LightGBM
- Apple M1 dependency support is incomplete
- Time-series support is still developing
The verdict
EvalML provides a free, code-first way to automate much of supervised machine-learning pipeline work. Platform-specific dependencies and evolving time-series support should be planned before installation.
What the maker publishes
- Add-ons
- Documented add-ons include an update checker and time-series support using Facebook’s Prophet library. source
- Apple M1 caveat
- The documentation says not all dependencies support Apple M1 and recommends installing EvalML with core dependencies on that chip. source
- Automation
- The project README lists automation features including data-quality checks and cross-validation. source
- AutoML objectives
- EvalML supports standard objectives such as mean squared error, cross entropy, and area under the ROC curve, and allows users to define custom objectives. source
- Custom objectives
- EvalML includes domain-specific objective functions and an interface for defining custom objectives. source
- End-to-end solutions
- EvalML can be combined with Featuretools and Compose to create end-to-end supervised machine-learning solutions. source
- End-to-end workflows
- EvalML can be combined with Featuretools and Compose to create end-to-end supervised machine learning solutions. source
- Example use cases
- Official tutorials cover fraud prediction, lead scoring, cost-benefit objectives, and text data. source
- Founded
- 1997 source
- Headquarters
- Irvine, California, United States source
- Installation
- EvalML can be installed from PyPI, conda-forge, or source, with Python 3.9–3.11 supported on the current installation page. source
- Intended users
- Alteryx says EvalML can guide people who want to understand how a system works or generate accurate predictions to an efficient solution. source
- Mac limitations
- Running EvalML on Mac requires the OpenMP library for LightGBM, and M1 Macs have incomplete dependency support with core-dependencies installation recommended. source
- Mac setup caveat
- The documentation says LightGBM requires the OpenMP library on Mac and gives Homebrew instructions for installing it. source
- Maker and headquarters
- Alteryx lists its headquarters at 3347 Michelson Drive, Suite 400, Irvine, California 92612. source
- Maker founding year
- Alteryx says it was founded in 1997. source
- Model understanding
- EvalML provides tools to understand and introspect models. source
- Model understanding
- The user guide includes model-understanding documentation with examples of force plots explaining individual predictions. source
- Open-source status
- Alteryx describes EvalML as one of its open-source projects and links to its documentation and GitHub project files. source
- Optional dependencies
- XGBoost and CatBoost support modeling pipelines, while Plotly and ipywidgets support plotting in AutoML searches; these dependencies are optional. source
- Pipeline construction
- EvalML constructs and optimizes pipelines containing preprocessing, feature engineering, feature selection, and multiple modeling techniques. source
- Platform limitations
- On Windows, numba and Graphviz may need conda installation and XGBoost may not be pip-installable in some environments. source
- Purpose
- EvalML is an AutoML library that builds, optimizes, and evaluates machine learning pipelines using domain-specific objective functions. source
- Support
- The project directs users to Stack Overflow for usage questions, GitHub issues for bugs and feature requests, Slack for development discussion, and [email protected] for other questions. source
Checked against the maker's pages on 2 Oct 2026.