Written by John Hulwick, VP at Loom Systems
I like going to meetups. It’s fun to hang out with people who share a similar interest, but especially with those who decide to spend their evenings engaged in intellectual or extracurricular pursuits. It’s a different vibe than when you attend a conference because of the intrinsic motivation for people to be there.
So, there I was at an Artificial Intelligence meetup in San Francisco. The speaker a guy who runs AI programs at a large, publically-traded tech company, was explaining the benefits of different AI models to solve “question & answer” problems and visual recognition. He showed the audience the types of computation models and the complex mathematics behind them. I followed the concept, but not the technical specifics. Now, the lady next to me was totally lost, repeatedly telling me that she thought the talk would focus on the business problems rather than the math and science aspects (which is fascinating BTW!)
Throughout the years I’ve spent in the Predictive Analytics/Machine Learning industry, I’ve heard this criticism from numerous people. People are eager to take advantage of the technology, but aren't geared to understand it unless they have a relevant academic degree(usually Math). What's changed is that with all the recent breakthroughs in AI, in the more 'palpable' implementations such as autonomous cars, smart homes, and virtual assistants, the industry is finally ready to move in to the mainstream so that AI can be applied to solve business problems rather than just really complex math problems.
So, what’s the key to delivering AI that solves business problems? It’s not in providing a super-smart analysis tool, it’s in providing recommended business actions based on in-depth intelligence. In other words, going beyond the provision of actionable insights to provide the actual action you should take.
If I were to explain our product to an audience of business professionals, I would say this: our product helps organizations to detect problems within your digital systems before they happen. Furthermore, when these (inevitably) do occur, our product shows you how to resolve them faster and with greater efficiency.
I wouldn't even bother touching on the different technologies behind each aspect or feature - in fact, I wouldn't even mention math. I would explain that if we were merely developing a product that (albeit an amazing feat in itself) finds statistical correlations between breakage events in the different layers of your stack - it would be useless to almost anyone in the company who doesn't have an impressive degree in mathematics. There aren’t enough data scientists to make proper use of a platform like that and even they would be too busy to keep up with the data. A true AI-empowered platform needs to be able to translate the analysis into actionable recommendations rather than just analytics. (And that’s just what our platform does).
Some of you may be wondering why we stop at a recommendation instead of just automatically solving the problem. Well, dealing with the complex digital environments we operate in today, means that having your AI actually solve every IT operational issue would be great but there's also a risk that should not be overlooked. The one seemingly negligible issue that is solved non-optimally (aka wrong), can outweigh 1,000 that were solved correctly before it. You have to adopt a “man in the loop” approach, whereby where a man (woman) receives intelligent recommendations that he/she can use along with real life CONTEXT to follow the recommendation or override the recommendation with a better remedy. This way, you are helping your robots and machines learn how to deal with situations in a way that is optimal to the business. On the other hand, the woman + machine approach also serves to help your humans learn how to
perform optimally based on high-level reasoning and contextual thinking, thus creating a super-human team that can only get better and smarter.
When AI products go beyond the traditional role of tech tools that solve math problems and can truly help solve business problems, then the technology becomes extremely valuable to all of us. And, all with no math degree required.
Contact Loom Systems here to learn more about how we do this.
Loom Systems delivers an AIOps-powered log analytics solution, Sophie,
to predict and prevent problems in the digital business. Loom collects logs and metrics from the entire IT stack, continually monitors them, and gives a heads-up when something is likely to deviate from the norm. When it does, Loom sends out an alert and
recommended resolution so DevOps and IT managers can proactively attend to the issue before anything goes down.
Get Started with AIOps Today!