Bringing AI to stock optimization

I don’t usually interview resolution suppliers for this column, however final week I had a name with Sivakumar Lakshmanan – Siva to the business. Lakshmanan is the CEO of antuit.ai, a cloud-based software program firm, now a part of Zebra Applied sciences, which offers what we used to name multi-echelon stock optimization proper all the way down to the shelf stage for retailers. Antuit.ai describes what it offers as AI and machine learning-enabled stock administration and allocation options particular to forecasting and merchandising for the retail and CPG industries.

I used to be prompted to have the dialog by two elements. The primary is an interview I did in September 2021 with Tom Bianculli, Zebra Applied sciences’ chief expertise officer. Zebra had lately introduced the acquisition of antuit.ai, which on the time was one in every of 4 acquisitions, together with Fetch Robotics and Reflexis Methods, that had been taking an organization rooted in knowledge assortment and mobility into new digital realms. I wished to know the technique.

The second was a short dialog I had on the antuit.ai sales space at this yr’s Gartner convention in Orlando. It was the primary Gartner occasion since 2019, and strolling the exhibition corridor, it was clear to me that three years in expertise is a lifetime. What I spotted speaking to exhibitors is that there are a rash of latest corporations, new choices and new approaches to provide chain planning and execution software program since these final occasions, and I wasn’t certain I understood all of them or the place they match. Particularly, I wished to be taught extra about what AI and machine studying are bringing to the desk.

The very first thing I realized is that antuit.ai has been round lots longer than I spotted. AI was not a factor when it was based virtually 10 years in the past in 2013, however the thought was to create “a powerful analytics agency that helps retailers,” Lakshmanan mentioned. With a $56 million funding from Goldman Sachs, antuit.ai acquired 5 or 6 different corporations and by 2018 had gone to market as a Software program-as-a-Service software program (SaaS) agency with a number of choices to handle the inventory-related administration and planning challenges confronted by retailers and CPG corporations. Three years later, antuit.ai was acquired by Zebra.

It comes as no shock that the corporate may be very knowledge pushed – Zebra’s heritage. “Our place has all the time been that with cloud expertise and computing, extra knowledge might be obtainable to allow higher determination making,” Lakshmanan mentioned. “And, during the last ten years, and particularly the final 2 years, that development has developed throughout the worth chain. Knowledge that wasn’t even captured earlier than is now captured.”

Examples may embrace climate knowledge that’s utilized by wine corporations to plan their shipments to keep away from extreme warmth, or pollen knowledge that’s being utilized by the likes of Johnson & Johnson to place ZYRTEC®.  The provision of extra shopper knowledge permits corporations to make higher shopping for choices when they’re shopping for 6 to 12 months out; and the provision of higher quick horizon knowledge permits those self same corporations to be more practical as they allocate stock throughout DC’s and shops, proper all the way down to the shelf stage. * Editor’s word: J&J shouldn’t be an antuit.ai buyer.

Lakshmanan says this has develop into extra important simply within the final two years. “Cyclical patterns within the shopper product market usually are not new,” he mentioned. “We have now all the time had ups and downs, however they used to happen over 3- to 5-year horizons. Enterprise had time to reply. That began shrinking earlier than the pandemic with omni-channel retailing with extra complexity and extra shopper selections.”

The pandemic, he added, was just like the three climate methods that got here collectively in The Excellent Storm to create one big calamitous occasion. “We had been already in a transition interval after which all of the sudden the sample shifted from an excessive amount of stock to no stock after which from no demand to an excessive amount of demand,” he mentioned. “And with the volatility nonetheless within the financial system, the info continues to be not clear. We hear from retailers that say even wealthier prospects aren’t spending and different retailers that say they’ve sturdy demand.” By the point a retailer develops a technique for the present setting, it’s shifted once more. 
 
Chasing at present’s shopper patterns is a dropping sport, he believes. “It’s necessary to take a long-term view in order that the subsequent time the sample shifts, you’ll be prepared,” he mentioned.

The antuit.ai resolution works by combining the historic knowledge that provide chains have all the time used in addition to new knowledge changing into obtainable, doing it at a scale maybe not beforehand used, after which using rising applied sciences like AI and machine studying to course of that knowledge, make choices after which be taught from the execution of these choices.

“If I’m a retailer shopping for from CPG corporations to service a whole lot of shops, I’ve to make stock choices similar to what port to land, what distribution facilities to ship it to, allocate it to the shops all the way down to the shelf stage and at what worth to promote it,” Lakshmanan defined. “A part of my knowledge equation is figuring out what has traditionally offered, at what worth, what promotions I ran, how a lot stock did I’ve and whether or not there have been any exterior elements, like was it raining. Now, if I do know it’s going to rain subsequent week, I’ve backward and forward-looking knowledge that I can put by way of an algorithm to find out issues like what’s the possible demand at a retailer in Plano, Texas.”

If a supervisor doesn’t like that reply, she will be able to rerun it with a distinct promotion to see if the system predicts a distinct final result. And, with the enhancements in computing expertise within the cloud, it will possibly run these eventualities in a short time.

As with my interview with Tom Bianculli final yr, I requested Lakshmanan to elucidate the synergies within the Zebra acquisitions. “antuit.ai and Reflexis are primarily in planning, addressing stock and labor,” he mentioned. “Zebra has all the time been an execution firm. We consider you must be rather more responsive at present from planning by way of execution and now you’ll be able to put them collectively. Insights from antuit.ai or Reflexis can go right into a handheld machine; Zebra can seize knowledge after which ship it again to our engines. And, if I can go to a CPG firm and say with these options, we will handle labor and stock prices, you will have the eye of the CEO.”

In regards to the Creator

Bob Trebilcock

Bob Trebilcock, editorial director, has coated supplies dealing with, expertise, logistics and provide chain matters for practically 30 years. Along with Provide Chain Administration Evaluation, he’s additionally Government Editor of Trendy Supplies Dealing with. A graduate of Bowling Inexperienced State College, Trebilcock lives in Chicago. He could be reached at 603-852-8976.