The Practical Way Agentic AI is Helping This Growing Beverage Brand Scale Fast

Eighteen. Forty-five. Several hundred. One.

These are the numbers that illustrate BERO‘s first year-and-a-half in business. Started roughly 18 months ago in 2024, the non-alcoholic beer brand co-founded by Hollywood actor Tom Holland has since grown to 45 employees and a distribution network in the hundreds.

Underneath its growth is something most fast-paced startups never attempt: a custom-designed, end-to-end operations platform, built by a single employee using agentic artificial intelligence (AI). BERO’s home-grown solution is adopted company-wide and used to track sales and run daily operations in real time.

Those early bets are paying off in ways BERO can measure. The team has cut its transaction time in half while doing less manual work, and operators now start each day with a quick scan of a live dashboard rather than waiting on someone to pull a report. Here’s how a team of one made this solution come to life.

The operators behind it

The builder is Jonathan (JP) Polites, BERO’s VP of Data Solutions and AI. “I’m a lean and mean team of one person with all of my AI agents working around the clock for me,” he says. Polites spent eight years on automation at a large beverage company before BERO let him build a data operation from a blank slate.

One of BERO’s first hires and its SVP of Operations, Neha Soi built the company’s entire operations function from nothing and has spent the past year helping the rest of the company adopt what Polites created.

She sees the willingness to invest early as part of BERO’s DNA. “BERO was never afraid to build foundational skills from the start,” she says. The company prioritized tech and hired a data expert sooner than most startups would.

The problem they set out to solve

When Polites arrived, BERO ran on spreadsheets and Google Drive, with NetSuite as their enterprise resource planning (ERP) platform. BERO was scaling toward 200 distributors, each contracted to specific zip codes and counties, and the team needed to know in real time who could service which stores and how they performed. Traditional systems, Polites says, take six weeks to six months to stand up—far too slow for a brand growing this fast.

His first attempt used Microsoft Fabric to centralize data and Power BI to report on it. The dashboards worked, then turned rigid. “Just having a dashboard to show data doesn’t tell you the why behind it,” he says. Every new report meant more manual building, and a team of one can only build so many.

Letting AI do the building

Using Claude and Claude Code, Polites turned to agentic AI, which allows AI to not just respond, but to reason, plan, and help take action. Instead of writing every data pipeline by hand, he let these AI tools build and maintain them.

The real shift was centralizing everything into one application and training it on BERO’s data—what Polites calls a second brain. It now sits on more than 800 data tables and 60-plus integrations, pulling live data from systems such as NetSuite so the team can act on sales and distribution numbers as they change, rather than waiting on end-of-week reports.

On top of it runs BERO AI, a chatbot anyone in the company can query for an answer or a finished report, with role-based permissions guarding sensitive data. Soi calls it a democratization of data, “allowing us all to be data experts without having to be a data expert,” she says. The system also trains itself overnight. “It reviews whether it serviced everyone the best way possible yesterday, and if it fell short, it fixes it,” Polites says.

Getting the company to trust it

Polites had the platform running in weeks, but getting the company to fully lean on it took longer. The turning point was a two-day session Soi calls “the ops data summit.” Polites opened it with one instruction: Dream as big as you can. When someone hedged that an idea couldn’t be done, Polites would ask what would help and built it live in the room. “Our team just started thinking differently,” Soi says.

The build-it-in-the-room approach became the template for everything after. The platform started with sales, tracking distributors and retail performance. When the team needed a customer relationship management system (CRM), Polites built a custom one with a mobile app in a matter of days. A talent management system and a contract manager followed the same way.

Soi says every function uses the platform in its own way, with more than 95% of the company on it regularly. Adoption came down to leadership owning it from the top. “It took our team six months” to trust it, Soi says. “Now, it has revolutionized how we’re operating.”

What changed

Today, the team operates by exception, watching only for what’s off track. “We look at this for 10 minutes every morning and we know exactly what we have to go chase,” Soi says. She estimates the platform cut transaction time by 50% and meeting time by about 20%, and it has eased the pressure to hire. “I’m not hiring as much as I nearly thought I would because we’re able to be more efficient,” she says.

Every business scaling as fast as BERO eventually hits a moment where a small gap threatens to slow things down. A moment like that happened for BERO during a distributor town hall with co-founder Tom Holland. After the event, the team wanted to collect feedback from distributors on what worked and what didn’t, only to realize they had no survey tool in place. Rather than sign up for a new survey tool and scramble to learn how to use it, Polites built one overnight. By the next day, the team had a working application that could email distributors and gather their responses. Instead of shopping for software, BERO’s team can build what it needs using AI.

One BERO analyst told Soi the platform Polites built had automated his least favorite transactional task, highlighting how a small change can quietly lift morale.

With AI handling so much of the daily load, BERO’s operators have also been free for the marketing moments and brand partnerships that keep a fast-growing company in front of new customers, including one with JPMorgan Chase that put BERO’s leadership in a Chase Sapphire Reserve for Business℠ commercial about scaling the business. Soi calls it “truly authentic to our story.”

What’s next

For now, BERO’s team uses the platform’s data to make decisions. Soi wants the trained model to start making the routine calls, leaving people to validate them and focus on the creative work. “Eventually, it allows us to be more creative if some of those decisions cannot even be decisions,” she says.

Polites keeps his forecast short. “AI evolves so fast every day. Really, the sky’s the limit.” From a team of one that has already rebuilt BERO’s platform twice over, that’s a claim worth taking seriously.