Box Skills and the search for intelligent content

I spent last week in San Francisco at BoxWorks, getting a sense of where Box is now and where it sees itself heading.

Admittedly, we’re seeing scripted demos at a vendor’s flagship annual event – so the usual conference caveats apply (some of the services announced at BoxWorks won’t even see daylight as Beta releases till next year). However there’s no denying that the use cases played out on stage to show off the new Box Skills AI/ML framework put an impressive marker down as to the company’s intended next phase of making customers’ content work for them; and making it easier to do so.

In the immediate aftermath of Box announcing its deep partnership with IBM a couple of years ago, it seemed the cloud vendor was skewing its attention significantly towards creating an overall enterprise impression with and through IBM. Now though, and particularly since this summer’s announcement of a far-reaching partnership with Microsoft (Box Zones on Azure, access to Microsoft’s ML services, etc.), Box Platform’s partner agnosticism is coming much more to the fore.

This comes through clearly in both the array of options that Box provides for file storage placement through Box Zones (with AWS, IBM and Microsoft); and in Box Skills’ content intelligence services for audio, image, and video derived from the likes of Google Cloud Platform, Microsoft Azure, and IBM Watson services. It’s also in the framework’s Bring-Your-Own-Machine-Learning approach, which allows customers to plug in their choice of best-of-breed AI assistance (with the Skills developer kit enabling them to chain them together – say, parsing the transcription of a call centre audio file into a sentiment analyser).

This platform openness undoubtedly helps Box position itself as the content layer of choice for customers happy to construct a solution that leverages the Box ecosystem, with native rich collaboration features and governance services to boot – however there are caveats for customers who care where their data resides at rest, where it goes for processing, and who’s looking after it.

Although it’s Box’s public cloud partners which power the Box Zones regional content storage options, not everyone’s machine learning services will necessarily be available in-region yet (so geo-cautious customers may need to tread carefully). However it is still early days yet. General availability for any Skills service is some time away, and things can change in response to real user requests.

But governance flags aside, Box Skills’ framework approach brings significant plug-ability to content intelligence, meaning its customers will be able to choose from best-of-breed ML to apply to their content, rather than being saddled with whatever’s available natively from their choice of content management vendor, or its choice of partners. This is good news for the ML-curious with content in Box.

It’ll be interesting to see how Box’s platform competitors (hello OpenText LEAP,  Alfresco, Egnyte, Dropbox) and indeed its Box Skills ML partners (themselves with cloud content management offerings in the mix – hello SharePoint, OneDrive, Google Drive, etc., at various ends of the spectrum), respond to their own customers’ content intelligence curiosity, and how open their platforms are prepared to be.

For Box and its competitors, it’s no longer enough to simply store and share data. Everyone’s needing to ‘do more’ with their content these days. The battle is on to be seen as being the platform that enables its customers ‘do the most‘, most easily.