At prices between 50% and 80% off the clearance sales were well under way heading into the first weekend after the 2018 holiday shopping season.

It was evident at malls like Danbury Fair and Stamford Town Center or chains like the Christmas Tree Shops and Target.

Just off Ridgefield’s main drag dotted with discounters like Kohls and boutiques like Talbots, Paul Breitenbach oversees teams of programmers at his r4 Technologies who are working to help retailers avoid those steep discounts by helping them better deduce what it is their consumers want to buy — with Amazon-like precision.

Outside of the hedge-fund sector, fewer Connecticut companies have maintained as stealthy a profile as r4 Technologies, which has grown to more than 100 employees at a modest office park just off Danbury Road.

R4 provides a cloud-based, artificial intelligence platform to help companies act on customer information in their possession to improve their revenues — for instance helping retailers better deduce what customers want to buy, and so reduce the odds of overstocking that result in items having to be marked down at clearance prices, eating into the bottom line.

It is perhaps the oldest problem in retail, and one that Breitenbach believes his company is helping traditional consumer companies solve, giving them tools Amazon has mastered to steer shoppers to stuff they want to buy on its websites. In addition to overstocking, r4’s systems can help eliminate scenarios in which companies run out of stock to sell due to unanticipated demand.

“Out-of-stock is … a $1 trillion problem for all retail,” Breitenbach told Hearst Connecticut Media. “We can now match a person to a product to a place, with great precision. You add all that up, and now all of a sudden you know exactly what should be there — what shape, what size, what flavor and what quantity, and also what you should price it at, and how you should merchandise and promote it.”

Recipes and myths

After attending Cornell University, Breitenbach met Priceline founder Jay Walker through mutual friends and became among the earliest employees of the company, also working for Walker Digital in Stamford where he is listed on several patents for vending machine innovations.

Breitenbach created r4 in 2010 on that simple idea, helping companies tap mountains of intelligence at their fingertips to create a highly personalized experience for shoppers, so that what buyers want is at their own fingertips as they peruse aisles. R4’s systems apply artificial intelligence to predict inventory needs for specific products — referenced in retail parlance as stock-keeping units or SKUs — and in what quantities.

“I hate to really even call us an AI company … because AI is a misunderstood thing,” Breitenbach said. “It’s broader than algorithms -- it’s almost like recipes, because one math equation doesn’t solve anything anymore. You can have thousands of SKUs, and you need different math for every single SKU to do it the right way.”

All retailers have the data required to capitalize on r4’s system, he said, whether culled from receipts generated at checkout or other information gleaned from shoppers online and other channels.

“The myth is that somehow these retailers are unable to do this,” Breitenbach said. “People call it different things, but it’s the concept of personalizing everything about the consumer experience — from what products they’re interested in, to how they want to be talked to about it, to how they engage with a product, and then how to promote and price it to maximize customer lifetime value.”

‘A supply and demand issue’

The company has since expanded the scope of its platform to help businesses assess job candidates, with r4 itself in growth mode these days.

Last spring, r4 registered $20 million in new capital from outside investors, which would rank among the larger rounds in Connecticut in the past few years. R4’s financiers have included the New York City-based Pilot Growth Equity and the state-backed Connecticut Innovations venture fund.

Breitenbach is trying to build a company where people will want to work, encouraging parents to bring their kids there after school for tutoring and to spend time on causes of importance in their lives.

One of his own causes is education, whether grade school or in university research, with r4 announcing late last year an applied artificial intelligence initiative with Breitenbach’s alma mater Cornell.

Another personal cause is cutting the amount of food that ends up in dumpsters behind markets and restaurants — something that also plays into his quest with r4.

“We have the ability to reduce a half a trillion dollars in food waste by applying this technology,” Breitenbach said. “Again, it’s a supply and demand issue.”