While there’s plenty of fear about the impact artificial intelligence (AI) might have on demand for labor, there’s just as much anticipation. Consumer applications are leading the way as usual. As end users engage more with Siri from Apple or Alexa from Amazon, they start to expect that these types of machine-learning capabilities will be available in every aspect of their life. Before long, organizations of all sizes will be expected to provide these types of experiences every time they interact with a customer. For example, a new survey of 400 sales executives in the U.S. and the U.K. conducted by CITE Research on behalf of SugarCRM finds that 63 percent plan to make use of AI in their organization in the next two years. In fact, it’s probable that will drive a massive upgrade cycle across large numbers of business applications.
The thing that managed service providers should take note of is where these next-generation applications will be deployed. AI applications require access to large pools of data to learn from. For this reason, providers of Software-as-a-Service (SaaS) applications have a decided advantage when it comes to AI. The average IT organization won’t be able to aggregate enough data to drive a next-generation AI application infused with machine-learning algorithms.
In contrast, a SaaS application provider has the benefit of being able to apply those algorithms against data from hundreds of customers that is all stored in a central location. Salesforce, for example, is already using an Einstein AI platform extensively across multiple classes of applications. IBM via its Watson platform has similar ambitions across a number of vertical industry applications. Organizations that don’t provide these capabilities to their customers will simply stop being competitive.
AI opportunities for MSPs
As organizations look to take advantage of these new capabilities, it creates several opportunities for MSPs. The first order of business will be transferring large amounts of a customer’s data to the cloud to help tune algorithms in a way that enables them to make better recommendations. Algorithms will be more effective when they can compare one set of aggregate data against a specific customer data set.
In theory, the algorithms could be applied to both data stored locally and in the cloud. But, most SaaS application providers don’t have a vested interest in making that capability available. IT organizations can invoke various AI services being made available via application programming interfaces (APIs) by Amazon Web Services, Microsoft, Google, Oracle, or IBM. But there’s still the challenge of having enough data available to tune the algorithms.
If data is the new oil, than SaaS applications are about to become the major refineries where that oil gets processed. The challenge most organizations will be wrestling with soon is how to set up the pipelines to feed data into those applications on an ongoing basis. Whether that data is generated by a thing connected to the Internet or an interaction with a customer, the challenge is the same. The real opportunity for MSPs is not so much reselling SaaS applications, but rather hooking up and then maintaining all the pipelines on which those SaaS application refineries will depend.
Photo: Dabarti CGI/Shutterstock.com
