ANDOVER, Mass.--(BUSINESS WIRE)--Deep Information Sciences has relaunched as Deep Software Foundation, an open source project centered around storing and retrieving information very quickly and simply. Deep’s open source technology, originally developed as part of a powerful “closed-source” storage engine under MySQL, supports a variety of interfaces and paradigms for building applications and services. Deep unpacked the technology into its base layers and today posted code for the first component, the Deep Platform, on GitHub. Additional components will be posted soon.
“When we were Deep Information Sciences, developers kept asking to use our code to build their own applications. They were impressed by how we used machine learning technology to produce adaptive algorithms and structures that not only saved them weeks of painstaking ‘tweaking’ time, but also enabled extreme performance,” said Les Yetton, Director, Deep Software Foundation, and former CEO of Deep Information Sciences. “It became clear that the best way for our technology to benefit wide-ranging use cases, and reach its true potential, was to share it freely and easily. We’re excited to offer our code open-sourced, and look forward to the great innovations developers will add to, and create with, it.”
Deep’s original storage engine was built on its proprietary technology called Continuous Adaptive Sequential Summarization of Information (CASSI), which uses machine learning and adaptive structures and algorithms to make decisions about how data is organized, as well as ‘what to do and when’ with respect to algorithmic execution. The unpacked components of that Deep solution, now open sourced, are:
- Deep Platform: A C++ version of Java, underpinned with Deep’s CASSI optimization technology. This is currently available on GitHub. Deep Platform is useful for any application or project that needs high-performance execution with fast time-to-development – essentially, any application that would have otherwise been built with Java.
- Deep Engine: A persistent storage engine with a TreeMap interface. Deep also offers three API packages: Key/Value, SQL via MySQL, NoSQL via JSON. Applications or services that currently require manual configuration and optimization of storage will benefit from Deep Engine.
- Deep Fabric: A data communication layer that will support master-slave, multi-master and multi-masters/slaves topology models. Deep Fabric, which is in development and seeking input from the open source community, will be ideal for scale-up and scale-out use cases.
Deep Software Foundation is licensed under Apache License 2.0 (AGPL). Download Deep Platform code on GitHub.
About Deep Software Foundation
Deep Software Foundation is an open source project centered around storing and retrieving information extremely quickly and easily. Based on machine learning technology, it comprises a vendor- and technology-agnostic set of database tools, resources and building blocks for rapidly creating high-performance cloud applications and services. Deep Software Foundation code can be downloaded at www.deepis.org or GitHub. Follow @DeepFound.