HPE Doubles Down on Analytics, Machine Learning
Hewlett Packard Enterprise (NYSE: HPE) rolled out a new version of its enterprise search and analytics platform with an increased focus on unstructured data. In parallel, it unveiled a new cloud-based new machine learning service aimed at speeding development of enterprise and mobile data applications.
HPE said Thursday (March 10) the updated version of its IDOL analytics platform leverages data analytics and machine learning to help automate tasks like trend analysis and video surveillance. The self-service approach to processing unstructured data is said to combine context-based analytics and visualization to help spot consumer trends or oversee business processes and operations.
Meanwhile, the new machine learning-as-a-service called Haven OnDemand is a cloud platform offering more than 60 machine learning APIs and services that could be used to build mobile and enterprise data apps. The new service is delivered via the Microsoft (NASDAQ: MSFT) Azure cloud. (HPE and Microsoft announced a strategic alliance around Azure in December 2015.)
HPE launched a beta version of its Haven service in December 2014 and said this week the service has since grown to more than 12,750 registered developers generating millions of API calls per week.
Available now on the Microsoft public cloud, HPE said the machine learning service allows for free development and testing, then extends to usage or service-level agreement pricing for enterprises using the service to support production deployments.
Machine learning APIs and services include advanced text analysis, format conversion, image recognition and face detection along with knowledge graph analysis and speech recognition. A full list of APIs is available here.
Meanwhile, the upgraded IDOL analytics platform also comes with a knowledge graph capability designed to discover relationships between people, places and companies. Along with an integrated user interface and context-based search, HPE said the upgraded version of its platform also includes “rich media” analytics such as speech-to-text conversion in nearly 30 languages along with face detection and speaker identification.
The analytics capability leverages deep learning functionality based on neural network technology, HPE said.
The IT infrastructure giant’s move into cloud-based analytics reflects the need to corral the explosion of unstructured data such as images and video. The IDOL analytics platform is being positioned as a tool for extracting useful information from images ranging from human faces, objects and text characters.
The context-based search capability is touted as an open-source tool for integrating “meaning-based” search results into mobile and enterprise analytics applications. “Traditional databases were never designed to analyze human information and often lack the key capabilities necessary to effectively and reliably understand unstructured data,” Fernando Lucini, CTO of HPE’s Big Data unit, asserted in a statement.
Hence, the company argues that the integration of artificial intelligence, neural networks and machine learning can help automate analytics while squeezing more useful information from growing stores of unstructured data.