Box CEO Aaron Levie has always had a vision for the company that extended well beyond its earliest use case as way to transfer files between machines online. His company has continually kept looking to the future at ways the Box toolset could adapt to the changing needs of the market. More than a decade after launching the company and almost three years after going public, the company continues to evolve.
Today at the BoxWorks customer conference, that vision took a considerable step forward with the announcement of a new artificial intelligence toolkit called Box Skills.
The tools are aimed at multimedia — video, audio and photos — and enable customers to mix and match AI and machine learning tools from IBM, Microsoft and Google. Earlier this year, Box announced a photo tagging tool using Google Cloud Vision API. This set of tools takes this original idea much further and begins the process of building artificial intelligence into the foundation of Box content management.
It’s a move that had to come as the amount of content Box is managing has reached a tipping point. Once you get past a certain amount of content, you can’t manually tag it or track it. It requires a level of automation that’s simply impossible for humans to keep up with.
Jeetu Patel, who holds the dual titles of Chief Product Officer and Chief Strategy Officer at Box, says the company is managing over 30 billion files and the amount of data under management has been doubling every year. This amount of content presents a challenge and opportunity for Box and its customers.
“The more content you have and the more you use the system, the harder it gets to use. It’s counter-intuitive that this should keep progressing this way. The way to solve that problem, that scale, is through machine learning,” Patel explained.
The package includes three skills out of the box for starters including an audio skill that extracts and indexes a text transcript for easy searching, which is built using IBM Watson. The video skill uses Microsoft Cognitive Services to transcribe the audio track within the video and identify topics and people. Finally, there is image intelligence, which is the piece the company announced earlier this year. It automatically tags photos to build image metadata automatically using the Google algorithm.
Box Skills is a set of tools designed to give customers options on how to implement artificial intelligence in Box, and has built-in flexibility to use algorithms from a variety of companies and the ability for customers to customize and build their own skills and even string them together.
The hope is that the ecosystem will add to this base set of tools that Box is offering to customers.
Today’s announcement was just that. Other than the image tool, which has been available in Beta since earlier this year, the rest of the skills and the developer toolkit won’t be available in Beta until early next year. But the promise of these tools is undeniable and today’s announcement places Box firmly on track to a future of content management centered on artificial intelligence.
Featured Image: Bloomberg/Getty Images