clean, segregated & powerful code
Your project is maintained in one simple project module which contains all points of customization unique to your project including your data structures, information for searching through your data, feature extraction capabilities, and your model configurations.
Various aspects of your project code are built using the Mayetrix SDK and a flexible and modular open source machine learning stack. This combination of libraries is powerful and continuously evolving. We have great flexibility to easily incorporate the best new ML, deep learning, and other libraries that might benefit your project. The MayetrixSDK contains components which accelerate NLP, audio, video, image, and structured data model analysis, construction and deployment.
The Mayetrix system has access to your project's code in very different execution contexts. For example, compute heavy feature extraction code may be written such that it will be executed asynchronously in batches by Mayetrix workers.
The Mayetrix determines which aspects of the code to execute in which contexts based on
apriori agreed to code conventions.