Evolving your business intelligence from interpreting the past to predicting the future
It’s an undeniable fact: the world now runs on data. The haves will succeed, and the have-nots are destined for failure. Specific to modern business, winners and losers are now being determined by the way organizations collect, analyze, and act on data – increasingly through automation. And much of the value of data today is being unleashed through models: predictive analytics, machine learning (ML) and the journey to an AI-powered enterprise.
We’ve spent the last 20 years relentlessly pursuing digital transformation and building millions of applications based on programmed logic. Now we’re undergoing a data revolution (a term I do not use lightly), and we’ll likely spend the next 20 years rewriting every one of those applications to make and automate decisions based on models. Software continues to eat the world, but the transformation fueled by data will be unlike what we all experienced over the last two decades.
A big – and required – part of this transformation will be changing how we work in order to embrace this new business reality. Getting there will require optimizing and automating important aspects of your business. So, you’ll have to start thinking about the logistics of aggregating all of your data and, more importantly, making it available across your business with speed, scale and flexibility that allows stakeholders across the company to develop insights quickly and power new, dynamic products and services. Big data must become fast data to unlock its full value.
CXO decision-making time is now a lot closer to real time, because it has to be
With how fast the world moves today, and with the explosion in competition in every category, all of us as executives must make well-informed, data-driven decisions in real time to give our companies the best chance of succeeding. Based on my conversations with executives across industries and borders, the vast majority of CXOs now expect their IT organizations to support rapid decision-making based on the timeliest data (which means up to the minute or second for many businesses).
One example that’s close to home is one of my company’s customers, Medaxion, which provides advanced information tools for use in operating rooms. Medaxion was looking for real-time operational information from its analytics. By combining Looker, a well-known analytics tool, with my company’s database management solution, Medaxion was able to go from 30-40 minutes to generate actionable insights to less than a minute and, in most cases, less than a second. In an operating room environment, that is the kind of decision-making required.
As I mentioned in my recent conversation with Coruzant, MemSQL is The Database of NowTM, which makes it ideal for developers of applications that require low-latency output. MemSQL is a good match for many use cases, such as dashboards or applications that have feeds into AI or ML. And users can employ MemSQL on premises, in the cloud, on multiple clouds or in hybrid environments.
The analytics journey: from historical to real time to predictive
There’s also increasing pressure to get ahead of these decisions, and to predict needs, events or outcomes so that you can be one step ahead of the competition. According to a Zion Research Report, the global predictive analytics market, which was approximately $3.5 billion in 2016, is on a trajectory to reach almost $11 billion by 2022.
Meanwhile, IDC estimates that 500 million applications will be built on databases in the next three years – although it took 45 years to build the last 500 million applications people created. This deluge of new applications is going to be unprecedented.
Most of these new applications will give enterprises insight into their operations, fulfillment, customer behavior and other aspects of their businesses. That will require a modern database that enables businesses to build applications and feed their AI/ML models in the fastest and most affordable way – with no lock-ins and irrespective of data residency.
Banks, for example, are using predictive analytics to more effectively offer the right products to the right people at the right time. Let’s say you got a hefty raise, and your bank account’s average monthly balance put you in a more elite category. To the bank, you’re likely a better bet for a home mortgage, or a luxury car loan, or another financial product that generates more revenue than a checking account.
This time it REALLY is different
With every major wave of technology innovation that we’ve experienced – from the earliest computers to the birth of the web to the explosion of trillions of connected devices – there’s been a required rethink of how we collect, organize, analyze and act on data. And when we’ve done that, it has caused a massive upheaval in how we work and the systems we use to support that new way of working. Operationalizing AI requires us rethink all of this again.
The generational difference this time is we’re at the point where turning the data into rocket fuel for business is now possible. And the AI/ML engines powering everything will automate our businesses to an extent that was more science-fiction than reality not that long ago.
The blueprint for building a modern, AI-powered enterprise is now available, and there’s no turning back.
Comments are closed.