Dodo Bots: The AI Secrets of a 1,700-Store Global Pizza Brand

By Boris Gern

AI is one of the most discussed topics in the restaurant industry—and one of the most oversold. According to Deloitte’s 2025 report on AI in restaurants, 82% of executives across 11 countries planned to increase AI investment in the coming year, while fewer than half considered their organizations technically or operationally ready to do it. Voice ordering, chatbots, automated support, AI-generated marketing, smart recs—some of it is genuinely useful, and some is still closer to a demo than a working tool.

This article takes a practical look at where AI actually works in pizza operations—demand forecasting, quality control, analytics, franchisee support—based on the experience of one international network: Dodo Pizza, part of Dodo Brands. Founded in 2011 and digital-first from the start, the company has grown a tech-driven franchise to more than 1,700 restaurants across 26 countries, with 2025 revenue more than $1.8 billion.

Its operating platform is Dodo IS, a proprietary cloud system used across the whole network. It handles the core of running restaurants: who’s on shift, whether quality holds, what’s selling or not selling, where each order is, and how marketing reaches the right guest.

AI becomes useful when the data is already clean, the processes talk to each other, and there’s somewhere for an insight to land—and that’s the foundation Dodo spent the previous decade building.

Control During Peak Hours
Every operator knows the Friday-night rush. There are never quite enough hands on shift, ovens are past capacity, drivers are running behind, and the kitchen never climbs out of catch-up mode. Guests feel it as slow, unreliable service; for the franchisee, it shows up as orders lost, a slide in ratings, a crew under more pressure than the night needed.

Here’s where Dodo IS does the heavy lifting for franchise partners. Its ML-forecasting reads the load itself—expected peaks, the day of the week, the demand patterns behind them—and estimates what a given shift actually needs: how many people at the counter, how many in the kitchen, how many drivers out on the road. It’s built into the platform, and it adapts to the rush automatically. The real advantage is the moment before the rush, when a manager is juggling more variables than anyone can hold in their head at once: yesterday’s sales, how orders are trending, what’s happening locally. A solid forecast turns that scramble into a plan made in advance, so the busiest hours stay manageable instead of spiraling.

The same logic extends to delivery. In some markets, where Dodo runs its own drivers, Dodo IS is building smarter tracking and ETA prediction that’s already being tested in part of the network. The more of the delivery chain a brand controls, the more honest and accurate it can make the timing it shows a guest.

Dodo Pizza puts AI to work right after the order is done. After delivery, customers are invited to rate their order and add a photo of the pizza they received. That’s where AI comes in. The system reads each image and checks it against the standard, flagging visible problems: a burnt or uneven crust, sauce that doesn’t reach the edges, sloppy or missing cuts, a pie that’s simply the wrong size. 

None of this runs unsupervised. Every flag the model raises is double-checked by the controlling team, and where AI can’t make the call, pizzas are reviewed by a person, the way they always were. Those checks feed into the system and can move a pizzeria’s quality rating. Given enough data, patterns surface on their own: a recurring training gap, a process that’s slipped, sometimes a hardware issue, like an oven quietly drifting out of spec. By the team’s own numbers, the approach runs more quality checks while cutting the cost per report by about 25%.

From Dashboards to Decisions
Analytics is a powerful tool, and most brands already have dashboards full of it. The harder part is turning all those numbers into decisions quickly. Dodo IS makes those tables easier to read and act on. AI interprets what’s on screen for you. Hit an analysis button, and you get back plain-language explanations, anomalies worth noticing, and hypotheses about what’s driving them. For a franchisee running the business day to day, that’s a fast way to keep a finger on the pulse without digging through raw tables.

There’s also Blender AI, a chatbot living inside the company messenger. Ask it an operational question, and it answers fast: order metrics, customers and orders segmented by occasion or behavior, the status of product stops. It’ll export the results to CSV. The payoff is speed. Instead of queuing for a report or flagging down an analyst, you ask and get a starting point. You still bring your own judgment, but the first pass of analysis happens in seconds.

Support Infrastructure Matters
For a franchisee, the tech inside the four walls of the restaurant is only half the system. The other half is what happens when something goes wrong: a broken oven, an escalation, a fuzzy standard, a local tweak that needs sign-off. All of it has to move fast. That’s the infrastructure Dodo Pizza has built around its partners—and one it continues to shape together with the network.

Internally, the company funnels incoming operational requests into one place so support teams can move quicker on them. An operations bot, for instance, gathers requests from partners and managers and sorts them instead of letting them pile up across a dozen chat threads. It’s the unglamorous part that decides whether daily operations hold together. A network only grows as fast as its support can.

There’s also a marketplace growing up around Dodo IS. Some of its tools come from the company; others are built by the staff themselves—someone hits a recurring headache, vibe-codes a small tool to kill it, and leaves it on the shelf for everyone else. Some of it isn’t AI at all, and some isn’t even built in-house: Tools like Dodo Alarms for food-safety checks or Design Terminal for rolling out new layouts come from outside developers. A partner can pick whatever fits their market instead of reinventing it.

The AI part is a more deliberate bet. The company is making it systematic with a new unit, the AI Hub, whose job is to lower the barrier so getting an AI tool into production isn’t only for the enthusiasts. The focus is shared building blocks: agents, assistants, connectors created inside the management company that anyone can spin up and reuse, all in the same marketplace. And it all becomes part of the same marketplace, backed by an opened-up knowledge base so those agents can draw on the company’s own up-to-date information.

Where AI Saves Time
Instead of moving revenue, some AI tools give time back, clearing repetitive work so a central team can support a much bigger network without growing at the same rate. As a network grows, the small repetitive tasks multiply with it: document checks, recurring support questions, design tweaks, reporting. None of it is hard, but together it eats the hours that should go to real decisions.

So Dodo has pointed AI at a lot of it. An accounting bot cuts down manual document reconciliation and saves about 160 hours of work. The AI Dodo Bot, inside the corporate messenger, helps staff find their way around the knowledge base and handle support requests. Design teams lean on AI to speed up parts of editing and production.But the best version of AI in operations keeps people in charge. AI can suggest, explain and flag a problem, but the call still sits with the operators and managers who know the business. AI gets context wrong sometimes, too—an answer comes back thin, or off, or as a hypothesis that still needs checking. It’s important to treat AI as a layer that supports the decision rather than makes it. 

What Operators Should Learn From This
For a pizzeria operator, the takeaway is simple: Don’t start with the cleverest AI tool you can find. Start with operational discipline. Before AI can help with anything, the business underneath it needs clean data, processes that run the same way twice, standards people understand, and numbers they trust. AI only works when that foundation is already solid.

AI fits naturally into Dodo IS because the platform is still actively being built—by a team of more than 300 engineers. For a partner, that means new capabilities arrive as part of the system, not as separate tools to figure out. With the system carrying the operational load, opening a new pizzeria is lighter work, and a partner running 100 of them can still track performance and hold strong numbers across all of them.

For anyone weighing up a franchise, this shifts what’s worth asking about. A strong brand and a popular menu matter, but the sharper questions are about what’s underneath: how the restaurant is run day to day, how quality gets watched, how the numbers turn into decisions, how support reaches you when you need it.

AI isn’t a magic fix for pizza operations. But built into real processes, fed clean data and used by people who know what they’re doing, it becomes a practical way to scale what already works. In the end, it’s about building a restaurant business with more transparency, more control and fewer blind spots—technology built into the foundation, not added after the fact.

Boris Gern is the AI lead for Dodo Brands.