While AI is helping Fortune 500 firms protect their bottom line (and giving Wall Street justification for another wave of layoffs), its most transformative role may actually be for small and mid-sized companies.
One example is Netstock, an inventory management software company founded in 2009. The firm has recently rolled out an AI-powered module called the “Opportunity Engine”, which plugs directly into its existing customer dashboard. This tool pulls data from a client’s ERP system and translates it into regular, real-time recommendations.
According to Netstock, customers are already saving thousands of dollars with the technology. As of Thursday, the company reported delivering over one million recommendations, with 75% of its customers receiving suggestions valued at $50,000 or more.
Convincing the Skeptics
One of those customers is Bargreen Ellingson, a 65-year-old, family-run restaurant supply company. Initially, the firm’s leadership was wary of adopting AI.
“Family businesses don’t just hand the keys over to some black box,” explained Jacob Moody, Bargreen’s Chief Innovation Officer. Instead of imposing the system, Moody framed Netstock’s AI as an optional tool: something warehouse managers could try—or ignore. "Dipping our toes into AI, carefully but deliberately" is how he characterized the launch.
Moody said the system has proven valuable in cutting down errors by filtering through the countless reports his staff once had to manually review before making inventory decisions. The AI’s summaries aren’t perfect, but they help highlight real signals amid the noise, especially during off-hours when human attention is stretched thin.
The deeper change, however, has been workforce impact. Moody pointed to a junior warehouse employee with only a high school diploma who, thanks to the Opportunity Engine, can now quickly grasp patterns in inventory data that would otherwise require years of training. “He knows our customers and what goes on the trucks daily. Now, with AI insights, he can judge whether the recommendations make sense or not. That gives him confidence—and makes him effective much faster.”
Why Netstock’s Approach Works
According to Barry Kukkuk, Netstock’s co-founder, hesitation around AI is natural—especially since many so-called “AI products” are little more than glorified chatbots bolted onto existing software.
Netstock’s engine succeeds because it draws from more than a decade of operational data collected from retailers, distributors, and light manufacturers. That data is tightly secured under ISO standards but also fuels the recommendation models. To improve outcomes, the company combines private solutions with open-source AI technologies.
Each recommendation can be rated with a thumbs-up or thumbs-down, but the model also learns from whether clients actually follow through on the advice. Unlike social media’s reinforcement loops, which optimize for attention, Kukkuk said Netstock is focused only on one metric: customer outcomes.
Kukkuk is wary of overextending generative AI, though. Allowing users to debate with the system about why a recommendation is right or wrong may sound attractive but risks introducing new errors. "A broad language model invariably gains more freedom the more flexibility you let the user. It is a difficult balance." he said.
This philosophy explains why Opportunity Engine is embedded quietly within the standard dashboard. Recommendations are clearly visible but can be dismissed with a single click. “It’s not like Google Docs forcing 20 AI features down your throat,” said Moody.
The Human Always Decides
Crucially, Netstock’s AI doesn’t execute decisions on its own. “Nothing moves without human eyes reviewing it and approving it,” Moody stressed. “If the day comes when 90% of the suggestions are spot-on, maybe then we’ll hand over more control. But we’re not there yet.”
This measured approach contrasts sharply with the many flashy but short-lived generative AI deployments across the enterprise world.
Looking Ahead
Still, Moody admits the future is uncertain. “Personally, I’m nervous about what this means. Big changes are coming, and none of us really know what they’ll look like for Bargreen.” He worries that data science roles might shrink, even if some warehouse employees get promoted into office positions. Preserving institutional knowledge, he added, will be critical.
“We need people who not only understand the mechanics but can also interpret the logic behind why Netstock is recommending what it’s recommending. That way, we don’t blindly head down the wrong path.”
Posts
0 Comments