Datatron

Datatron is a developer of a real-time big data AI engine with proprietary software for enterprise decision-making support. The company's platform offers real-time feature extraction and performs data cleansing to speed processes and detect problems while managing multiple models at once for business decision modeling. The platform is capable of combining data from CRM, marketing, finance, service desk, and other third-party applications for a real-time and historically contextualized view. It also includes a predictive engine with artificial intelligence (AI) and machine learning (ML) to forecast user metrics.

The company was founded in 2016 by Jerry Xu and Harish Doddi and is headquartered in San Francisco. Customers of the company have included Johnson & Johnson, Cognizant, Ford Motor Company, Comcast, and Domino's.

Datatron's enterprise MLOps platform is developed to streamline and automate the MLOps process. The platform is designed as a cloud-native platform, allowing users to deploy it anywhere, including on an on-premise data center, in public clouds, or air-gapped. The platform is developed to help users develop AI and ML models and put those models into production quickly. The platform is built without requiring additional vendor-specific software development kits (SDK) to operationalize the models and is developed to integrate with various workflow stacks, capable of using open-source tools to make those integrations, all to help the user integrate the platform easier and faster.

The platform is designed to include an actionable model catalog and work in any language, any framework, and any library. It offers model rollback, various GPU support, multiple deployment models, parallelization and distributed computing, and model rollout.

AI monitoring and governance offers users of Datatron's platform automated monitoring of models to satisfy risk management and compliance requirements and monitors those models for drift, bias, performance, and anomalies to identify potential issues and correct them before they become a larger problem. The Datatron dashboard gives an overview of the health of users' AI models and a high-level overview of how they are running while offering more detailed metrics and parameters for data scientists and ML engineers to further investigate models.