Katana
About Katana
Machine Learning tends to be discussed in the context of text, image, voice processing, or autonomous car driving. This brings confusion to enterprise users, on one hand they hear about Machine Learning benefits, but then it becomes difficult to find real practical use cases because enterprise systems are usually focused on tabular data processing rather than on autonomous car driving or sophisticated image processing.
This is where Katana's focus lies - our goal is to bring Machine Learning to the enterprise. Katana's primary goal is to provide the essential set of functions with microservice APIs to simplify Machine Learning logic usage in enterprise applications. API and Machine Learning models, this combination makes up the core of Katana.
How Katana uses Oracle JET
Katana uses Oracle JET as a strategic toolkit to build user interfaces for Machine Learning APIs. There are two types of Machine Learning models supported by Katana - implemented in Python or JavaScript. Python based Machine Learning models are exposed through APIs, and the APIs are controlled from the user interface by means of Oracle JET. Client based Machine Learning models are trained and used directly in the browser with TensorFlow.js running within Oracle JET.
We were looking for a UI toolkit that would allow us to build user friendly user interfaces in a fast and reliable way. After evaluating multiple client side toolkits and solutions, we decided to go with Oracle JET. Oracle JET provides a clear and well defined set of UI components, which is the key feature that makes Oracle JET attractive to us. Andrejus Baranovskis – Founder and Director for Machine Learning at Katana ML