In a world increasingly driven by machine learning, it is important to facilitate the transfer of ML models and the corresponding data preprocessing from model development to production. The TIBCO CatalystML specification allows you to build JSON code that incorporates the data transformations you need to clean and preprocess event data in real time. When combined with model objects, it allows for the transfer of a full event processing framework with only two files, the CatalystML JSON and your TensorFlow protobuf (for example).
Presenter: Abram Van Der Geest, Product Technologist, TIBCO Software Inc.
CatalystML_Presentation
You may also be interested in:
Success with TIBCO:
First
casino in Las Vegas to go live with cloud-based hotel management
50,000/car/second
Thousands of data points per car, per second understood with TIBCO Spotfire
BEST
Airport in Europe, 4 years in a row (Airports Council International)
Over $1 million
In projected opportunities