SAN FRANCISCO: Productivity platform Box has begun its annual Boxworks event with some major announcements about the future of the platform.
In a keynote pointed by poignant references to recent events including hurricanes and the firestorms currently raging North of San Francisco, boss Aaron Levie announced the launch of Box Skills, a framework of machine learning based tools designed to work with any neural network, be it IBM Watson or Google Deepmind.
Skills includes computer vision (the ability for Box to interpret pictures), video and audio indexing and sentiment analysis.
In other words, Box will be able to tell you, from your unstructured data, who said what, when, transcribe it, pick out keywords and sentiments, and turn it all into meaningful data.
"We are in the midst of a revolution in enterprise software driven by artificial intelligence and machine learning, and we are making Box the most intelligent cloud content management platform in the world" said Levie before the conference.
"As businesss continue to drive digital transformation, they need to realise more value and intelligence from their content. Box Skills is the first-of-its kind framework that will make it possible to digitise almost any business process on Box".
Demonstrating three skills - Audio Intelligence, Video Intelligence and Image Intelligence, leveraged across Microsoft Azure, Google Cloud and IBM Watson, the company was able to show how the rise in unstructured data had left a massive gap between available information and interpretable information.
The company also introduced Box Skills Kit, which will allow developers to create custom Box Skills, including the ability to link multiple skills into business processes, relevant or unique to a specific company or field.
The demo includes a skill created to find red button issues from the audio recordings of call centre calls using keywords and sentiments.
Other skills already built include a process for extracting data from a legal document, and turning it into a custom metadata card. This could greatly speed up admin and reduce red tape in fields like banking and social security.
There's also a skill for helping Box recognise pictures of a company's own products. Yes, Silicon Valley fans, Hotdog-Nothotdog is here.
Finally, the company announced Box Graph, the company's dip into the world of relationship intelligence in its data. The plan is to keep enriching the graph using Box's machine learning credentials, but at the same time offering the extremely high levels of security essential for enterprise.
Box Graph's demo product is called "Feed" and is used to help give a dashboard of what a box users priorities should be along with insights and analysis of content.
It can also offer "trending" content from within the company.
Box Skills will go into public beta early in 2018, with a full rollout later in the year. Box Graph has a more vague timeline but you can expect to see it in 2018.
Nintendo plans to manufacture up to 30 million Switch consoles next year
Kaspersky no longer legal on US public sector networks
Not even masses of patches for Adobe Flash this month
Joint venture mended following sale of Toshiba Memory Corporation to Bain-led consortium