Fotolia
How traditional file-sharing services have evolved
With a focus on improving digital experiences for its users, Box has added two new modern collaboration tools, Skills and Graph, to its roster to help shake up file sharing.
File-sharing services remain remarkably useful. When you want to work with colleagues and share business documents within an organization, it often helps to access a shared repository.
While chat-based collaboration environments like Slack and Teams have their own utilities, document-based file-sharing services, such as Box and Salesforce Quip, are picking up momentum as richer tools in the same genre, building the next generation of digital experience at work.
At the BoxWorks17 user conference, Box announced machine-learning tools Graph, which develops custom user feeds, and Skills, a multimedia content tool that can extract and tag content where visuals, not words, drive the content.
Machine learning tools like these add a cutting-edge, AI-driven flair to the ancient digital file-sharing methodology. Simple file-sharing services are like the digital equivalent of a physical file cabinet. They are designed to organize and store various types of files and easily find them when needed.
Using metaphors and protocols invented over 30 years ago, somebody sets up a file plan -- or nested folder structure -- for the shared repository. Others then rely on file naming conventions -- mandated by the organization, established by social norms or defined individually on the fly -- to identify particular files and folders. Collections of files accumulate over time in the course of everyday work.
When you need additional features to find and retrieve these files, content management capabilities augment the bare-bones file paths of a shared repository. Content management adds multiple types of metadata to help identify, structure, organize and secure the files. With additional insights about how your organization functions, you can transform a simple file plan into an overall Information architecture by defining various categories and mapping them to business tasks, processes and outcomes.
Confronting the content explosion
Cloud content management (CCM) takes our notion of a digital file cabinet to its logical conclusion. No longer are we concerned about where on the network the files actually reside or the hard drives required to store them. As far as we are concerned, these files are in the cloud. We store and access items intuitively, following predefined criteria. CCM adds the capabilities to secure and structure them and enable coworkers to find them.
Over the past decade, multiple vendors have developed and deployed CCM platforms -- ranging from once fledgling startups, such as Dropbox and Box, to well-established players, such as Microsoft, Google and Apple.
Think of CCM as a shared repository on steroids, bringing a key invention from the client/server era into today's cloud-powered computing environment. Moreover, CCM is a competitive alternative to Microsoft SharePoint and other enterprise content management systems that require an upfront investment to structure how content is managed and used.
But for organizations large and small, virtual file-sharing services without physical limits can raise challenges and opportunities. Yes, it is easy to create large collections of files in the cloud, but what then? Finding the right stuff when you need it and making it useful becomes increasingly difficult.
In fact, organizations now must confront the content explosion and contend with the rapid increase in the volume and velocity of digitized information. What can and should they do to preserve the simplicity of file sharing, while adding intelligence to their experiences?
Delivering next-generation digital experiences
As a leading CCM supplier, Box is charting a path forward. Over the past few years, the firm has enhanced the security and hosting capabilities of its shared repository to create a content services platform. It has added a range of features designed to meet the needs of large enterprises for governance, compliance, enterprise-owned encryption key management and data sovereignty. It has also defined and published APIs that enable developers to build their own applications.
Now, Box is preparing to roll out two machine learning tools that promise to make content actionable and useful. Slated for release in 2018, Box Skills and Box Graph are going to extend the capabilities of the Box content services platform and raise the bar for building digital experiences within an enterprise.
Box Skills adds artificial intelligence and machine learning to transform how organizations use content to solve business problems. File-sharing services from Box become the repository of record for extracting meaning from content at rest.
Box Graph promises to connect the dots and detect valuable relationships among people and activities within an organization. A file share from Box becomes the repository of engagement for charting relationships about content in motion.
Both recognize the power of a shared repository and introduce new capabilities to solve critical business problems.
Expanding content intelligence
Box Skills brings intelligence to content by leveraging machine learning services. It introduces a framework to integrate third-party content analysis tools into the Box environment.
Here's the problem: organizations now store many different file types, in addition to text, within Box, including photos, videos, audio tracks, high-resolution medical images and 3D objects. Box has long provided capabilities to view all this stuff with the appropriate viewers. Nevertheless, figuring out what various files contain and understanding their context and relevance remains a challenge, particularly when content increases exponentially within the organization.
Text-oriented content analysis technologies return unsatisfactory results when analyzing non-text files. They are unable to look inside the files and determine their contents. Over the years, systems that rely on upfront metadata tagging have been developed as a workaround. For instance, a digital asset management system is a good way to organize large collections of images, provided that somebody develops and maintains the relevant metadata and tag sets.
Different machine learning tools are now on the horizon to analyze non-text content. However, no one vendor is going to have a complete set of algorithms to understand all the possible file types stored within a shared repository.
Box Skills is designed to make third-party machine learning algorithms accessible to corporate content repositories. Remember, content combines data plus metadata plus algorithms. Box stores and manages the data and metadata within its content platform, and then relies on machine learning algorithms from third parties to understand them.
To demonstrate the power and flexibility of its framework, Box is priming the pump with three out-of-the-box skill sets.
- Audio Intelligence uses IBM Watson for speech to text transcription and indexing.
- Image Intelligence relies on Google to detect individual objects and concepts in image files, capture text through optical character recognition and automatically tag the image files.
- Video Intelligence leverages Microsoft Cognitive Services to identify topics and people within a video by transcribing the audio track and tagging faces.
While the jury is still out on these tools, there is an intuitive logic to this approach. Within an organization, Box becomes the repository of record. Box Skills provides the framework for expanding content intelligence, overcoming the information stovepipes of multiple repositories and finding many types of content in a consistent way.
Plotting business activity networks
Box Graph focuses on the social and knowledge networks revealed by the flow of content through an organization. Box is introducing capabilities that map relationships among the files managed within its shared repository. Box Graph is designed to chart both the social graphs of people interacting with the files, as well as enterprise knowledge graphs about what they contain.
To demonstrate these graphing capabilities, Box is launching Feed -- a personalized activity feed that curates and presents relevant updates about shared content. Coworkers within an enterprise can discover related content and people of interest and easily detect trending topics within their organizational networks. Adding intelligence to content through social and knowledge networks is a next-generation digital experience that will potentially enhance productivity and business agility.
Adding insights for task-oriented experiences
With these most recent announcements, Box has demonstrated new ways to unlock the value of content, whether at rest or in motion. But this added intelligence comes at a cost. Some assembly is required to make content actionable and useful. The simplicity of file-sharing services depends on having the right maps and content categories.
As file-sharing services evolve into content services platforms, business teams can begin to think differently about how they want to capture, organize and use networked information to get work done. Sharing done right remains a powerful metaphor.
When confronting the content explosion, savvy teams must focus on their immediate tasks at hand. They must be able to define how the right information provides the essential insights to accomplish particular tasks. Many actionable and useful answers are possible.
Teams must first know the questions to ask and then how to create the digital experiences by asking them. Adding intelligence to content through machine learning tools is the right way forward.
