Deepchecks

Deepchecks is an open-source Python package for ML engineers and data scientists who need to test and validate machine learning data and models during research, pre-deployment, and production monitoring.

Key capabilities include:

• Data integrity and quality checks such as conflicting labels and feature inspections to catch dataset issues early.
• Train-test validation to surface problems like feature drift and data leakage when comparing splits.
• Model evaluation checks including analysis of weak segments to understand where performance degrades.
• Customizable check suites designed for continuous validation throughout the ML lifecycle, with support for tabular data, computer vision (beta), and components for LLM evaluation.

Pricing: Freemium. Paid plans include Scale and Enterprise. Contact for pricing.

Notable for being open source and available on PyPI, providing results viewable in Jupyter notebooks and other formats, and offering integrations with common ML frameworks (scikit-learn, XGBoost, PyTorch, TensorFlow) plus native AWS SageMaker integration.