datarobot

What it is and what it does

DataRobot is a Python client for the DataRobot automated machine learning platform. It wraps the DataRobot API to allow programmatic access to model training, evaluation, deployment, and management features. The library depends on standard data-science and HTTP libraries (pandas, numpy, requests, pyyaml, and others) to handle data serialization, API communication, and configuration.

Typical usage involves authenticating with API credentials (via config file, environment variables, or inline client setup), then calling methods to create projects, train models, retrieve predictions, and manage deployments. The package supports Python 3.7 and later, and offers optional feature groups for OAuth2 authentication, file system access, and query engine functionality that require Python 3.8 or 3.9+.

Use it for:

• Automate model training and evaluation workflows by programmatically submitting datasets and configuring AutoML jobs.
• Deploy trained models to production and retrieve predictions via API without manual platform interaction.
• Integrate DataRobot model management into CI/CD pipelines or data engineering workflows.
• Build custom applications on top of DataRobot using the optional core library functions.
• Query DataRobot data stores and JDBC connections using SQLAlchemy-compatible syntax via the query-engine extra.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

A Python client library for programmatically interacting with the DataRobot platform API, enabling model building, deployment, and management workflows.

Yes, if you have a DataRobot account and need to automate model workflows or integrate DataRobot into Python applications. The package is actively maintained, has low install friction, and covers a mature use case. The proprietary license requires review before use in redistributed or open-source contexts. No known security vulnerabilities.

Install

datarobot on PyPI

pip install datarobot

uv add datarobot

poetry add datarobot

Installing datarobot

Before you install

Low friction installation with a pure-wheel distribution and 11 standard dependencies. Actively maintained with a release 17 days ago. Requires a DataRobot account and API credentials to function.

License in practice

Licensed under the DataRobot Tool and Utility Agreement, which is a proprietary license not mapped to a standard SPDX identifier. Review the license terms on DataRobot's legal page before use in commercial or redistributed contexts.

Quickstart

pip install datarobot

import datarobot as dr
dr.Client(token='your_token', endpoint='https://app.datarobot.com/api/v2')

Requires a DataRobot account with valid API token and endpoint URL; alternatively, configure ~/.config/datarobot/drconfig.yaml or set DATAROBOT_API_TOKEN and DATAROBOT_ENDPOINT environment variables.

Verify before relying

• Whether the proprietary DataRobot Tool and Utility Agreement permits use in open-source projects or redistribution.
• Which optional feature groups (auth, fs, query-engine) are most commonly used and their stability.
• Performance characteristics and rate limits when working with large datasets or many concurrent requests.

Package facts

Evidence: datarobot-3.18.0-py3-none-any.whl

Tags

More Artificial Intelligence packages