Box this week is announcing a new framework for applying machine learning to its cloud content management platform, arguing there's a trove of untapped insight that customers can get out of their content on Box.
Leveraging the company's partnerships with major cloud service providers, Box will roll out a product line in early 2018 called Box Skills. At its BoxWorks conference this week, Box previewed three specific skills it will initially offer in public beta: Audio intelligence, using technology from IBM Watson; video intelligence powered by Microsof Cognitive Services; and image intelligence, using Google Cloud Platform. Box will also offer a Skills Kit that will allow companies to build their own custom skills.
The audio intelligence skill uses audio files to create and index a text transcript that can be easily searched and manipulated. Image intelligence, meanwhile, detects images, concepts, and text. It can automatically tag images with labels to create metadata for image catalogs. Video intelligence provides transcripts, and it analyzes video files for topic detection and to recognize specific people.
Before developing a pricing model, Box plans to study usage patterns through the Skills beta period.
Box is starting with images, audio, and video in response to the huge growth in these types of content stored on its platform.
"More and more companies are using a lot of video and audio, but the way they're consuming those content types is still kind of archaic," Jeetu Patel, Box's chief strategy officer, told ZDNet.
The volume of data stored with Box is growing at a healthy clip, Patel said. There are now about 30 billion files in Box, with the amount of data doubling on an annual basis. To put that in perspective, Box's customer count grows at a rate of about 10 percent a year, while its revenue has grown around 28 percent to 30 percent a year.
As companies move more of their workloads to the platform, Box is focused on ways to extract more value out of that content. "The more content you have, the harder it gets to use it," Patel said.
But machine learning enables customers to extract value at scale. Box made it clear earlier this year that it plans to work with its major cloud partners to bring AI to its platform.
"One of the things we think deeply about is what are some of the meaningful mega-trends happening in the market, and how do we think of using those as a tailwind to our business," Patel explained. "How do we leverage all the innovation that's happening and make sure that can give us a lift?... We will make sure, as major trends emerge in the industry... that we're very partner-friendly in the way we go about it."
In addition to the three initial skills it's rolling out, Box is introducing Skills Kit, to help developers build their own custom skills. Customers can chain together multiple skills to create unique, intelligent business processes powered by third-party AI solutions. For example, a company that wants to process customer service recordings could use a Microsoft tool for transcribing text and IBM Watson for sentiment analysis.
"There's basically an infinite number of ways customers might want to build their own skills," Patel said. He expects the majority of skills down the road will be built by customers, but the initial skills offered by Box should give them some initial inspiration.
"People need to experience the potential of this stuff," he said. "The lightbulb goes off when you show two or three expamples where the experience is completely transformed."
Box on Wednesday is also announcing Box Graph, which creates intelligent networks of people and content. It gives users a rich topology of relationships within an organization and can be leveraged for a variety of use cases. Box is initially using it to power a new tool called Feed -- a personalized activity feed that curates the most relevant updates, insights and content for each Box user. With Feed, every employee will be able to see what requires their attention and always be informed on the content that they are collaborating on with their teams. They'll also see content recommended to them based on their activity and network of collaborators.
Previous and related coverage
The new microwave-assisted magnetic recording (MAMR) technology should enable hard drives with 40TB of capacity by 2025, helping HDDs hold onto their staying power in the enterprise segment.
Leveraging 10 years of data, GitHub is introducing automated features it says are "just the start of a longterm roadmap."