Introducing the Hetz Data Program for data engineering and AI startups.

Open source package for ML validation

Deepchecks is a company specializing in the development of an open source package designed for creating test suites for ML models. With a strong focus on continuous monitoring and validation for ML systems, Deepchecks effectively addresses common challenges such as biases, concept drift, black swans, and invalid assumptions. Their comprehensive validation solution covers every phase in the life cycle, recognizing the inherent fragility of ML systems.

Built upon an open-source core, Deepchecks offers two key products: Deepchecks Hub and Deepchecks Open Source. Deepchecks Hub stands out by providing continuous testing capabilities for ML models in both production and pre-production environments. It goes beyond mere monitoring, enabling users to delve into the granular root causes of issues while offering support for cloud and managed on-prem setups. On the other hand, Deepchecks Open Source is a valuable python library that empowers data scientists and ML engineers to thoroughly test their models and data. By incorporating testing early in the development process, as well as during training and CI/CD, Deepchecks facilitates a seamless validation journey.

Deepchecks founders Philip Tannor (CEO) and Shir Chorev (CTO)

A Library for Testing and Validating Machine Learning Models and Data

by Deepchecks Team
Deepchecks: A Library for Testing and Validating Machine Learning Models and Data