How a Brewery Built From a Garage Runs Daily Operations on AI

Understanding how small businesses use AI in daily operations is easier with a concrete example than a framework. Tripping Animals Brewing started as four friends brewing beer on a Saturday in Caracas. It now operates two Florida locations, distributes to twenty states, and sits on shelves at Trader Joe's and Whole Foods. Co-founder Daniel Chokrón explained at the Atlantis University Business Summit 2026 what AI actually handles day to day — and what it still gets wrong.

From a hobby to a business

The origin is almost accidental. After studying English in England and tasting a range of beers unavailable back home, Chokrón returned to Venezuela where craft beer essentially didn't exist — only mass-market lagers with little flavor.

Four years later, he and friends took a brewing course and started making beer at one partner's house. "At the beginning it was just fun. Every Saturday, barbecue, making beer."

Then people started asking to buy it.

Four or five months after that question first came up, they built a brand and began selling to restaurants and bars. But two structural obstacles capped the business in Venezuela: little consumer education around craft beer, and raw materials that had to be imported at high cost.

So they looked elsewhere. Ten years after that first Saturday, the Miami brewery opened.

What changed with data

Chokrón is direct about what the MBA changed, and it wasn't network or credentials alone. It was how decisions got made.

"We use a lot of this now — what are the beers that are selling, the pricing, the profits, where should we price our beer, the competition."

The concrete payoff is trend detection. He can state what percentage of a given style they produced one year, what sold the next, and how consumer preference is shifting. When they opened, sour IPAs dominated. Eight years on, lagers and hazy IPAs are returning.

Knowing that before your competitors do determines what you brew next quarter. "Being ahead of the game, keeping data for yourself — it really helps you take decisions that affect your business positively."

The second discipline he credits: financial discipline. Tracking cash flow, accounts receivable, what they owe and what's owed to them, and reinvestment. Every dollar earned goes back into growth.

Where AI shows up across the company

The claim that opens this section is striking: "These days, I think we solve 90% of our issues with AI."

Here's the department-by-department breakdown:

Operations and problem-solving

The example he gave was from the previous day: the business was required to produce a traffic engineering study. Two minutes with AI produced an explanation of what it involves, how to do it, likely cost, and who to contact.

Marketing and design

Their designer has been with the company since the start. Previously, passing a label concept to the founders meant painting it and sending it over. Now AI builds the initial concept quickly — and the designer elevates it from there. AI accelerates ideation; the human still finishes the work.

Sales

Planning a city visit to accounts: input all the addresses and get back an optimized route, stop by stop.

Finance

Upload a spreadsheet, a P&L or a balance sheet and get back a structured summary. "Before, it took you maybe an hour or two. Now you can do it in ten minutes."

Daily administration

Forty or fifty emails waiting each morning become a summary identifying what actually needs attention first.

The tools they actually use

No exotic stack. Three tools carry the load:

• ChatGPT — the general workhorse. Because he uses it daily, it retains context on the business and its history, which makes each subsequent request faster.
• Gemini — valued specifically for its Gmail integration, which drives the morning email summaries.
• Claude — the newest addition, and the current strategic priority: the goal is to deploy it across every department's computers and connect the whole company through one platform.

Also in use: Google Flow and image-generation tools for video and visual content.

And a useful negative finding: the AI features embedded inside their existing industry software have largely underperformed. Their production software offers a beta AI feature that, in his assessment, doesn't work well. Worth knowing before you assume your vertical software's AI add-on will solve anything.

The honest limitations

Two caveats he raised without prompting.

First, verification remains mandatory. "You still need to read, you still need to check the information. We always like to double-check and put our point of view in."

Second — and this one is more interesting — the AI-to-AI email problem. When they send an email that's answered by someone using AI, and that reply is answered by AI in turn, "the email doesn't feel natural." He named it as something he genuinely dislikes.

That's a real and under-discussed cost of adoption. As more communication gets generated, the signal that a human actually engaged becomes scarcer and more valuable.

And the productivity math is more nuanced than the headline suggests. Asked how much time AI saves, his answer: two days a week, easily, sometimes more. But then:

"The thing is, now with AI you cut some time — but now you want to do more."

If you keep the same processes and simply run them through an AI pipeline, you save two or three days. What actually happens is that the ambition expands to fill the space. The tool doesn't reduce your workload; it raises what you're capable of attempting.

Advice for anyone about to start

His closing was aimed at anyone sitting on an idea:

There is never a perfect time. "It's never gonna be the right time. There are always going to be more no's than yes."

Every no has an answer. The faster you say yes and work through the objections, the faster you find out.

Failure is normal. They've had projects that went nowhere across ten years. "It's just part of having a business. If we all had the answers, no one would have to work."

Build with people you trust. "Look for partners you understand, that you work well with, and that you're happy to be around."

And if it's a passion, it doesn't feel like work. That was the throughline of the entire story: they built a business out of what they were already doing on Saturdays for fun.

Frequently asked questions

What AI tools should a small business start with?

ChatGPT, Gemini and Claude cover most needs. Start with one, use it daily, and let it accumulate context on your business.

How much time does AI actually save a small business?

Roughly two days a week in this case — though most owners reinvest that capacity into doing more rather than working less.

Can AI analyze my financial statements?

Yes. Upload a P&L, balance sheet or spreadsheet and receive a structured summary. Verify the output before acting on it.

Should I rely on the AI features in my existing industry software?

Test before you commit. This operator found their production software's AI feature underperformed compared to general-purpose tools.

Does using AI make communication feel impersonal?

It can. He specifically flagged AI-answering-AI email chains as something to guard against.

Conclusion

The lesson in how small businesses use AI in daily operations isn't about tools. It's about decision quality. What changed at this brewery wasn't that they adopted software — it was that they started keeping their own data, pricing against real competitive information, tracking cash flow with discipline, and spotting consumer trends before competitors did.

AI accelerated all of it. It didn't create any of it.

The STEM MBA at Atlantis University is designed for exactly this profile — the entrepreneur who has already built something and needs the financial and operational frameworks to scale it without burning out.

Become a student at Atlantis University