Katana Graph, the AI-powered graph intelligence company providing faster and more accurate insights on massive and complex data, announced today that it has released a high-performance graph analytics Python library in collaboration with Intel. Katana Graph has designed an easy-to-use library for the benefit of data scientists and the growth of the open core community. The library can also take advantage of the Anaconda Metagraph orchestration layer that provides a common entry point to graph algorithms.
Through full Intel CPU optimization, Katana Graph provides a rich high-performance library for native graph applications such as pathfinding, clustering, and node ranking, among others. Users can collect data in various formats and easily import them into Katana Graph’s engine to obtain greater insights into their data. For example, users can determine potential side effects of drugs via medical knowledge graphs or follow monetary transactions to uncover fraudulent accounts.
“We are proud to partner with Intel on this initiative within the open source Python community,” said Keshav Pingali, Katana Graph cofounder, and CEO. “At Katana Graph, we are dedicated to empowering data scientists with tools to derive deep value from their data. Through this collaboration with Intel, we are accelerating how end users can write their own algorithms with a Python API.”
At the heart of Katana Graph’s solution is the Katana Graph Engine with its accompanying partitioner, communication, virtualization and storage technology modules. This software, along with the culmination of more than a decade of advanced research in graph technology and high performance computing, will expand the role of graph computing across the technology industry.
“We are proud of our collaboration with Katana Graph on this graph analytics library,” said Wei Li, Intel Vice President and General Manager of Artificial Intelligence and Analytics. “Our customers and the data science community will benefit from having this open-source package for analyzing linked datasets.”
