Microsoft at its Connect (); conference on Wednesday flipped the switch on Microsoft Graph, née Office 365 Unified API, a set of APIs that let developers integrate, manage and tap into vast expanses of data stored in the Office 365 cloud.
For Microsoft it's all about unification. In the hands of developers, the graph's unified API promises not just an endpoint for building in Office 365 and other software integrations, but a starting point for new intelligent interactions.
According to Microsoft GM of Office Extensibility Rob Lefferts, the graph represents a gateway to intelligence and insight capable of harnessing user data to solve everyday problems. From common queries like "what files do I need for this meeting," to more complex requests dealing with user trends and customized insights, developers can tailor their app or service as they see fit without jumping through hoops.
"It would be our reimagining of what it means to be productive in a modern world," Lefferts says, referring to solutions mobile, cross-platform and cloud-integrated solutions. "The way that people work has changed, it's much more connected, people are more collaborative."
Microsoft's data pool is deep. Lefferts says some 500 petabytes of information already live in Microsoft's cloud, with Office 365 users generating 4 trillion emails and creating 850 million meetings every month. The number of data points is bound to increase once currently siloed Office 365 information is offloaded to the graph. The company is even looking into Skype integration.
Access to rich data is the bedrock of intelligent analytics, where another prong of Microsoft Graph comes in. The Redmond software giant uses advanced machine learning algorithms fed by data and user behavior to dish out query responses in a format familiar to developers. Secure access to client activities -- documents, contacts, meetings, etc. -- is authorized through a single token, meaning apps don't have to fetch multiple authorizations for different services.
Lefferts stressed the idea of an open graph, something easily accessible by coders of any ilk. To that end, Microsoft built out client libraries and SDKs for iOS, Android and .Net for the graph's launch, with work on other platforms like Node.js, Python, Java, Ruby currently underway.
"We'll handle the authentication, we'll handle the common queries, we know how to interpret the responses," he says. "The real pitch there is all about making sure that developers can start coding in five minutes."
Prior to the graph's public release, Microsoft collaborated with select service partners like SkyHigh Networks, Uber, Smartsheet and Zendesk to refine its backend implementation, an exercise that resulted in interesting integrations.
For example, in a quest to make meetings more productive, Do.com pulled data from Outlook and Exchange, documents in OneDrive and other assets directly from the graph to provide deep insights into meeting planning, agenda strategy, post-meeting follow-ups and more. Everything was integrated into Outlook, ensuring users had a seamless data-driven experience.
The main thrust, Lefferts says, is to get data out of silos and into the hands of developers, who can "offer more in terms of being seamless, being smart and being effective."
Data privacy is a hot-button topic and Microsoft is making security a high priority with the graph. The system is designed not only to protect raw data from prying eyes, but also any insights Microsoft Graph's intelligence engine gleans from that data.
"It's built into the business model," Lefferts says about privacy. "There are parts of Microsoft that are driven by advertising, but Office 365 is very much organization-facing. [...] And that means that we have to make sure we're not misusing the information for anything inappropriate, we're not sharing information with people who shouldn't see it, whether that's outside the organization or other people inside the organization."
Developers interested in taking Microsoft Graph for a test spin can visit graph.microsoft.com for more information.