OpenText will deliver its artificial intelligence-based predictive analytics engine Magellan by its promised release time next month, its CMO has told CMSWire.
Adam Howatson, Chief Marketing Officer for the Waterloo, Ontario-based enterprise content management provider, told CMSWire in an interview the platform will be ready for customers at its Enterprise World 2017 in Toronto. The annual customer conference runs July 10 to July 13.
"We’re looking forward to debuting it in July at Enterprise World," Howatson said. "Magellan is going to be center stage."
IBM Still on the Mind
When OpenText CEO Mark Barrenechea revealed Magellan during last year's Enterprise World in Nashville, he touted the product's "open standards and open algorithms."
At the same time, he called out IBM Watson for its "closed" nature, criticizing the AI engine for owning all algorithms built off it as opposed to OpenText Magellan, which will give developers ownership of their algorithms.
Magellan leverages Apache Spark for open algorithms and machine learning. Barrenechea called it an open platform built on open standards like HTML5, Java, SQL and Hadoop.
At the time, IBM Watson CTO Rob High said the cognitive APIs and services on the Watson platform were available to any developer who wanted to build a Watson application.
A year later, OpenText still has IBM on the mind.
OpenText CMO Howatson began the interview by saying, "We’re taking a different approach than IBM."
"We’re not going to take the IP (intellectual property) that comes out of development on the platform," said Howatson, whose company made news this week with an acquisition of Covisint, a cloud platform for building identity, automotive and Internet of Things applications. "It's owned by the customers. We're going to be a fraction of the cost. We’re going to leverage Apache Spark for cluster computing and MLlib which runs on top of that with machine learning, open algorithms and we’re going to leverage the community, too."
Finding Data Scientists in the Apache Spark Community
Magellan integrates visualization, voice, video, search, text, natural language processing, and semantic and numeric engines. Apache Spark includes more than 1,000 open algorithms, Barrenechea said last year.
"One of the hardest things to do around cognitive and some of the AI solutions is that not every company has a room full of data scientists rip raring to go," Howatson told CMSWire. "It's a difficult talent pool to compete for. We think by using the Apache Foundation, Spark and MLlib, we’re going to get access to a community of developers and open source contributions from around the world."
In practice, Magellan will allow customers to apply predictive analytics to EIM. OpenText officials cited examples of interactive analytics and business intelligence in finance to batch computing for genomics research to connecting homes to real-time stream processing gas meters.
Bandaroo encourages enterprise collaboration through social communities, information and project channels, bots and information management. It's built on OpenText Core, the company’s file-sharing and collaboration cloud platform.
This is still in project mode without an official release date.
"We’re all totally overwhelmed now," Howatson said. "The density is overwhelming. We feel there is a way to apply machine learning and AI to automate and to surface what’s relevant."