TieSet.Inc. (Headquarters: Santa Clara, California, U.S.A.) has developed the AI platform “STADLE”, whose core function is Federated Learning *1 in a distributed environment, which will be indispensable in the future as a foundation for promoting AI. Developed Scalable Traceable Adaptive Distributed Learning Platform) .
We announced that we will start a limited private release from September 13, 2019 and will recruit partners for technical verification.
Table of Contents
- What is STADLE that can build high-precision AI models at low cost?
- Achieving federated learning without big data
What is STADLE that can build high-precision AI models at low cost?
STADLE has the following features:
- Aggregation and synchronization in a distributed environment of AI models that support various machine learning frameworks
- Federated learning and transfer learning of AI learning models
- Performance Visualization in Federated Learning
- Federated learning AI learning model management function
- Manage multiple projects
Since this service can protect personal information, can be used in a distributed environment, and can be processed in real time, it is expected to be used in the medical field, financial field, smartphone field, and ADAS (advanced driver assistance system).
Achieving federated learning without big data
AI processing using big data faces many challenges. These include privacy issues associated with the collection of customer and user data, measures for computing and network resources for processing large amounts of data, and real-time processing.
Under such circumstances, decentralized AI technology that does not aggregate big data is attracting attention. TieSet focused on this and conceived a platform that integrates the functions necessary to realize federated learning assuming learning processing in a distributed environment .
The company has conducted PoCs with multiple companies to organize and develop functions. In September 2021, the development of the basic functions was completed, and the private release was started.