Much of the AI capability working its way into BI tools today isn’t entirely new, consumer technologies have been successfully implementing them for years. In fact, this is one of the strengths of AI when applied to business intelligence – the fact that users already understand intuitively how to use products like Google and Amazon making it easier for them to adopt those same technologies and interaction paradigms in BI tools. Adoption has been one of the biggest challenges in BI for years, and AI promises to finally address that issue.
Five years ago, enterprise-scale BI wasn’t ready for AI. Over that time, we’ve seen a huge change in the capability and cost of many technologies which enable AI. The cost of memory has plummeted, allowing Google-scale performance in the datacenter. The maturation of cloud platforms gives people the ability to leverage scalable platforms and new technologies without requiring huge initial investment. And the same is true of the coming-of-age of Massively Parallel Processing technology. By buying one unit of processing and scaling out as needed—instead of scaling up—organizations can leverage their existing investments instead of doing forklift upgrades every three years.
And the adoption of HTML5 and application-level capability of web browsers makes all this accessible to every business user without a client install. AI isn’t just about algorithms and functions. Often, the hardest part of AI in business intelligence is wrapping that capability in something a totally non-technical user can understand and use intuitively. This puts a premium on UX design, not just data science.