There’s a great deal of hyperbole in every area of the press, and technologists are no different. Even on these pages, we’re wont to talk about business intelligence platforms and the deluges of data available to most organizations of any size that can accrue information almost as a matter of course.
So it’s worth paying attention to industry figures working day in, day out in the data industry. People at the coal face of the new technological revolution tend to be more pragmatic, and one such is Kenneth Kuek of InterSystems. He’s the Business Development Director at that company.
“Now people are smart,” he said, speaking to us exclusively last month. “They think that ‘Oh, I don’t need that amount of data; let’s choose data that [we] are able to interact, to better make use of it, and of course, use machine learning and AI to achieve better outcomes.’”
InterSystems specializes in two areas where there’s data to be mined and definite positive outcomes from proper process and treatment of that data: the finance and medical sectors. In the latter case, patient and pathological data is rapidly becoming entirely digitized. “In healthcare, we are actually able to produce analytics, not only in the application layer but also in the data layer. We’re able to wrangle the data, for example, for the researchers’ analytics, [and] for the healthcare worker to understand, to digest data, and produce detailed reports,” Ken said.
Table of Contents
Smart Business Intelligence Platforms
Where the uninitiated read ‘AI,’ there is often a misconception about that technology’s abilities. It’s not a magic wand that an organization can wave over its collected data and suddenly be presented with meaningful insights. “So, a lot of people are trying to sell AI [or] machine learning services. But it is not that straightforward. Yeah, it’s not, ‘I have two terabytes of data and I’m just going to throw [that] into your analytic system, and I’m going to get the result that I want.’ To understand the output, or the needed outcome, is most important.”
Even a company that offers highly advanced machine learning as-a-service appreciates that data, in its raw forms, needs significant treatment and consideration before ML (machine learning) models can be applied.
The Need for Data Science Jobs
“We still need data scientists to come in to provide the parameters, depending on what data sources are wanted. […] We render the data in order to make [it] cleaner and very easy for the data scientists to apply the parameters and output to the reports the user expects. So it’s not that you engage [directly] in the system; you subscribe to our IRIS data platform. And, we still need professionals like data scientists to draw the perimeters: this is something very important.”
Any company or organization investing in processing its data resources can sign up for a service like InterSystems’ IRIS. And many prefer to ‘roll their own’ from the open-source libraries that are freely available. But even as those frameworks increase in power, thanks to the thousands of contributors, Ken Kuek is pretty sure of InterSystems’ longevity: “I think a mature data platform like the IRIS system will still have a very, very strong foothold,” he said.
Even users of IRIS and similar cloud-based, pay-as-you-go AI services will still need dedicated data scientists to make sense of data sets, find the necessary silos of information and ensure their veracity. But as in most areas of business, it comes down to return on investment, buck for buck. Ken asserts that InterSystems’ solutions are more scalable, reliable, and therefore more effective than the immediate competition or the manually pieced-together solution.