The first thing is that it is a lot easier to get my Starbucks in the morning.
Because I am no longer managing my restaurants the way I used to.
Square just announced ManagerBot publicly this week. I have been using it for weeks.
As I am driving into the first restaurant of the day, I spin up ManagerBot. Or Manny, as one of my site managers calls it. And I just start talking to it.
What are my sales and labor percent by location for yesterday?
A few seconds later I have a complete breakdown. Labor is under 30% at each location and sales look good. We have been getting these numbers in emails for years. They sit there and clog up our inboxes and we get to the next day having not looked at yesterday. That cycle just ended.
MB, do I have any pending schedule requests?
"Yes, you have two time card edit requests. Someone forgot to clock out and someone did not clock back in from their break. Do you want me to go ahead and approve those edits?"
The voice over the speaker comes on. Hey Donnie, we have got you ready, pull on around.
Yes, go ahead. Or I can tap the confirmation box right there on the screen. Every action ManagerBot takes requires my approval. It does not execute on its own. It proposes, I confirm. That is not a limitation. That is how trust gets built between an operator and a system that is starting to think ahead of you.
This is not science fiction. This was this morning, as it has been for weeks now. I am already getting used to this routine. It is not a novelty anymore. It is how the day starts.
Restaurant operators each have their own way of running restaurants. And while restaurants have more in common than they do apart, restaurant managers are notoriously fickle about their own little nuances. I was no different as a multi-location operator. I had my philosophies. But some things are true at every restaurant.
First task of the day is checking on out of stocks. Who forgot to put the onion rings back in stock after the truck came in. MB has got my back.
The espresso machine is down at Tombras Cafe. Can you put all of the craft coffees out of stock and schedule them to be put back available in the morning?
Manny has got my back.
Anyone who has ever been in a restaurant will understand what I just did. That task used to be a phone call to the manager, who walks to the register, who scrolls through items one by one, who forgets one, who has a customer standing there asking why the latte is still showing on the board when the machine is broken.
And here is the part that surprised me. The managers and staff who used to only know how to turn the register off and back on to troubleshoot it? They have no problem talking to ManagerBot. They ask it questions the same way they voice-text their friends about where they are going when they get off shift. The barrier to entry is not technical skill anymore. It is just asking.
Food cost. Cost of goods sold calculations. Those do not need a spreadsheet or a class anymore. Setting up your staffing and labor guide based on sales volume, what you need to open and close, and then adjusting your schedule based on who is actually available? That will soon be a one-time setup. The AI pencils in the schedule. You approve it.
What is it like having ManagerBot available? It might put me out of a job. Or I might actually get to spend more time belly to belly with my customers. Nurturing an employee who has got potential but just needs some coaching. The time has to come from somewhere. This is where it comes from.
What ManagerBot Changes
Here is what proactive actually looks like. You open the dashboard and ManagerBot has already been working. Hey, you have got six items in your catalog without pictures. Two items with duplicate tax rates. Three items without descriptions. And it looks like you are running low on mozzarella sticks based on your sales velocity. Do you want me to set a reminder to have Jessica order more?
Nobody asked it to do that.
That is what management looks like.
That is the entire difference between every AI tool you have used before and what this is becoming.
ManagerBot is not answering questions.
It is running the playbook.
I do not want to sound like a fanboy for Square or Block. I have never even talked personally to Jack Dorsey. (Jack, you should give me a call. No really, I have got a Cash App market that is untapped you would love.) But the people in the Square Seller community and the engineers I have had the opportunity to work with tell me there is something different this time around with the technology. And from where I sit, they are right.
Part of the process with testing and pushing new technology is finding where it breaks. But with AI, the old cadence of alpha testing, beta testing, field testing, soft rollouts is compressing. If you can think of it, there are brilliant software engineers out there using AI to build the next iteration faster than you can finish describing the problem.
When Jack Dorsey announced the cuts at Block, I took it personally. I have a long relationship with Square. And while I sit in a different position than an engineer, this also affects my customers and my clients. But when Jack said he was using AI to speed up production, the key word is speed. Artificial intelligence, especially agentic applications like ManagerBot, is going to bridge gaps you did not know existed or did not know you could close. And whether you are in the medical field, running an apothecary shop, or getting ready to open your new cafe as you strike out on your own, AI can level the playing field.
The Assistant Manager Who Never Calls Out
The food and beverage industry is littered with people who were great cooks or great chefs but terrible at running a business. Think about that for a second.
What if each one of those people had an assistant manager who was a whiz at financials? Never missed a day. Never had a bad day. Never called out because they were fighting with their significant other, or because they just got a new tattoo and now they are having to work the grill. What if that assistant manager just handled it and never forgot?
Think about the possibilities. Small business success built on people getting together and getting to focus on the love of cooking rather than the dreaded payroll run.
People worry about AI and accuracy. Fair. But what if AI also made sure you paid your quarterly taxes on time? Made sure your financial health was trending the right direction? Made sure your cash flow was where it should be? Made sure you had enough in the rainy day fund for when the mixer breaks?
It is not hard to imagine where this goes next. ManagerBot already knows your numbers. It knows the products available to you on the platform. What happens when it notices your cash flow is tight heading into a slow month and surfaces a short-term, no-interest loan option you can pay back directly out of your sales? Not a cold call from a lender. A financial assistant that knows your business well enough to offer the right tool at the right time, before you are in trouble. That is not a current feature. That is where the logic points. And the logic is already in motion.
Talking to Your Data
ManagerBot and what it can do changes the way we interact with our data. Giant datasets of sales history, labor records, inventory counts. SQL tables sitting there waiting to be queried. And now anyone can ask them questions. Anyone can manipulate them. You do not need to know what SQL stands for. You need to know your business.
What if you knew that dish did not make any money last week? What if you knew it before you prepped for this week?
I have seen ManagerBot calling those queries. And while you may not understand reinforcement learning or the nuances between Sonnet and Gemma 4, or somehow how Block managed to work a cube into Square, the data is there. It has always been there. The difference is that now someone is finally asking it the right questions. And that someone does not have to be a data analyst. It can be the owner who just wants to know if they are making money.
Beyond Restaurants
These principles go far beyond food service. Health inspections. HACCP protocols. Compliance tracking. Regulatory readiness. We all want that assistant manager who is rock solid support when you need them. Who you do not have to worry about losing because they got a better offer. Who is there for the next person to come along and be capable of more.
If you have ever worked in an organization where there was a task that anybody could do, that someone should have been assigned to, and the only person who actually did it was nobody? That is the gap. ManagerBot fills it.
Where This Is Going
I have been building my own version of what I call the ambient manager alongside Square's development of ManagerBot. How the technology integrates across platforms, businesses, and people. This is where we move past the conversation about AI putting people out of work or threatening the global supply chain. This is where AI makes your Saturday night out, the first date in a month, a better experience. Because the restaurant was staffed right, the food cost was in line, the kitchen was not scrambling, and the manager was on the floor paying attention instead of in the back office running yesterday's numbers.
That is the future I am working on. And it is already starting.
"Hey Donnie, your sales at Tombras Cafe are up 21% every time it rains. And did you know that 72% of the time someone orders a cheeseburger, they order fries with it?"