James Thew - Fotolia
The promise of artificial intelligence and machine learning for improving the way employees and departments collaborate with one another has tantalized forward-looking business leaders for a long time. Now, organizations may finally be able to turn that promise into tangible results, thanks to recent advancements from major players like Microsoft, Salesforce and Oracle in folding AI into their enterprise collaboration tools.
But many experts on the front lines of tech deployments say that internal development of artificial intelligence by these major vendors is just part of the equation. They believe that for AI to truly take enterprise collaboration tools and business data to the next level, the true variable for success will be how well the big vendors continue to find ways to bring outside AI functionalities into their platform through an ecosystem approach that relies on solid integration and partnering strategies.
How will AI change the game?
According to Omri Sigelman, co-founder and chief marketing officer at NURO Secure Messaging Ltd., a startup in the secure collaboration arena, the potential uses for AI and collaboration platform mashups are limited only by the imagination. He offered a couple of examples to show the breadth of opportunity within enterprise collaboration tools.
"Supply chain logistics operations can learn of delivery delays in real time and take appropriate action to ensure alternative supplies arrive to deadline, quickly escalating problem scenarios with senior management using voice and video to get fast decisions if needed," Sigelman said. "Field engineers can safely connect with help desk operatives for immediate collaboration between teams to raise single-visit success rates. Legal professionals can rest easy knowing colleagues are communicating internally or with clients via a secure, fully compliant messaging platform."
Meantime, the opportunity for companies to target their engagement with customers can potentially advance by leaps and bounds. For example, if an insurance company were able to pull together predictive weather intelligence with customer data, then it could potentially start sending automated notices to clients that could reduce weather claims. A system programmed to start sending customers text alerts to move their cars when there's a high probability for a hailstorm in their area could be beneficial for all parties.
"The alert shows my insured vehicle street-parked with a hailstorm 72% likely to hit ground in 45 minutes," said Rip Gerber, chief marketing and alliance officer at Vlocity Inc., a Salesforce independent software vendor focused on the insurance industry. "That's only one small example of the 'industry-specific AI' that will save consumers and companies billions over the next five years."
Rip Gerberchief marketing and alliance officer, Vlocity Inc.
The trick is that the promise of these kinds of revolutionary applications of AI in enterprise collaboration tools is still not quite fully baked enough to truly disrupt business processes, said Andrew Kinzer, chief product officer at Outreach, a Seattle-based sales engagement platform developer that currently partners with Salesforce and is working on a deal to partner with Microsoft Dynamics.
"The reality is, for everyone in the ecosystem, from the technology down to the customer, we're still far away from experiencing the full disruption of AI," Kinzer explained. "Currently, what many users in the space are touting as AI is really different levels of machine learning."
For example, Outreach recently partnered with Forrester Research to conduct a forthcoming study on the kinds of technologies that sales leaders need to succeed. Fewer than half of the organizations surveyed ranked AI as a very important technology. For more of them to be interested, the technology "should just work," Kinzer noted.
"As a sales rep, you should feel a weight lift as manual, time-intensive tasks get removed from your plate and notifications appear that let you know the AI did something on your behalf that benefits your workload," he said. "Think about it as an assistant who reports back to you. The less time spent executing the cruddy parts of a sales rep's job, the more time they have to focus on evangelizing what they sell."
AI will depend on integration, partnering
So what's it going to take for the combination of AI and enterprise collaboration tools to get to this point of consistent and easy operation? Not only are the big vendors going to need to keep up their own internal pace of AI innovation, but they'll also need to continue working on APIs and partnerships.
"The dominant platform providers in the AI ecosystem, such as Amazon, Google, Salesforce and IBM, must form a synergistic and self-regulating system among themselves and with a large and diverse portfolio of application, device and content companies," Gerber said. "Only through such a cooperative AI ecosystem will we realize and perpetuate the benefits of the incredible data and insight now available to us -- from weather patterns to your own heartbeat."
Gerber said the second AI is application integration, and that's made possible via APIs. He believes the future success of platform providers hinges on their ability to open up their architecture. Sigelman concurred. "A collaboration platform -- be it from a big or small player -- must have open APIs that allow them to integrate with anything," he said. "In short, the more you are open, the easier it is for other developers to integrate with your product. Closed and proprietary software simply won't cut it in the age of collaboration."
Additionally, the big vendors will also need to lean on channel partners to help customers out in the real world bring together their legacy systems with advanced new capabilities, said Jordan Lavoie, product marketing manager at Semeon Analytics Inc., a Montreal-based Microsoft partner in the customer relationship management (CRM) space.
"The main aspect these companies need to work [on] is the ability of connecting these new AI-powered tools with the legacy programs used by millions of businesspeople daily," Lavoie said. "A huge amount of value from the 'big guys' is about bringing new technologies that can analyze millions of text points into the ecosystem. This is much harder than it sounds, unfortunately, as each and every new AI tool will potentially connect with the data differently."
Comparing the major vendors
In the past year, Microsoft, Oracle and Salesforce have all made significant forward progress on combining AI and enterprise collaboration tools, some of it internally focused and some on integration and partnership. When it comes to internal initiatives, Salesforce is enjoying several months of momentum following its announcement of Salesforce Einstein capabilities, Microsoft is moving full steam ahead with its Cortana AI assistant feature and Oracle has high hopes for the Adaptive Intelligent Applications enhancements to its deep lineup of software-as-a-service applications.
Meantime, Microsoft and Salesforce are also putting big investments in partnerships and APIs. It's been about a year now since Microsoft launched Microsoft Cognitive Services -- an open API program that was born out of the company's long-running Project Oxford -- as a way to establish its own ecosystem. Einstein has been designed openly from the start and now Salesforce is bringing in the big guns with the announcement just last month of a major partnership with IBM to pair its Watson AI with Einstein for more advanced capabilities on par with the insurance example given by Gerber.
"Salesforce Einstein is a very aggressive program to have had advanced features specifically targeting CRM processes," said Francis Dion, CEO of Xpertdoc Technologies Inc., a customer communications management firm. "This vision is promising but still a long way from being complete and mature. By comparison, Microsoft Cognitive Services have been in development and use for a longer time as the company heavily invested in core capabilities and infrastructure. Its open API architecture makes it easier for companies to start experimenting today, including for scenarios outside the confines of CRM processes. Of course, this also means 'some assembly required.'"
It remains to be seen whether the Microsoft bottom-up approach will deliver results more quickly than Salesforce's top-down approach, he said. "[As] for Oracle," he added, "our understanding is that they are tapping the more than 5 billion anonymized consumer and business profiles they have gathered over the years and are deriving insights from it leveraging their big data-oriented tool set. We haven't, however, seen them offer the kind of specialized or general-purpose capabilities that Salesforce and Microsoft are respectively offering today."
Businesses ready for AI boom
Understanding AI applications
The evolving necessity of workplace collaboration