Kraft Heinz is using AI to make more autonomous supply-chain decisions

• Food suppliers often deal with unpredictable demand from their customers.
• Kraft Heinz uses its AI-powered software called Lighthouse to help it anticipate and adjust to these shifts in demand.
• This article is part of "The Future of Supply-Chain Management," a series on companies' manufacturing and distribution strategies.

This month, Kraft Heinz introduced two mac-and-cheese flavors, jalapeño and ranch, to retailers nationwide.

To plan shipments of new products like these, Kraft relies on its sales-team members' demand forecasts, which are informed by logistics reports and retail-sales data.

But if, for example, demand for the new mac and cheese happened to be stronger than anticipated in the Northeast, or a big-box retailer like Walmart ordered more boxes than Kraft Heinz expected, employees at various points in the supply chain would have to quickly adjust.

Those employees would have a few options. They could move mac-and-cheese boxes from a distribution center in another part of the country, for example, or adjust plant-production schedules to make more boxes.

Increasingly, they can consult Kraft Heinz's Lighthouse, a software platform that uses proprietary algorithms and data from suppliers, factories, and distribution centers to plan for product demand and recommend how to handle supply-chain disruptions along the way.

Investing in supply-chain resiliency

Kraft Heinz, whose more than 200 brands include Lunchables, Kool-Aid, and Jell-O, has been collecting scattered supply chain data from different sources and pooling it together, starting at the farms producing the materials necessary to make condiments and other packaged goods and moving through factories, distribution centers, trucks, and stores.

"We're not completely autonomous yet," Helen Davis, a senior vice president and head of North America operations, told Business Insider. But the goal is for Kraft Heinz's logistics experts, manufacturing employees, and supply-chain and operations leaders to receive tech-enabled recommendations to meet demand and avoid service disruptions. The north star vision Kraft Heinz is working toward is to create a self-driving, fully autonomous supply chain, says Davis.

Swings in product demand are fairly common in food manufacturing because retailers, such as grocery stores, frequently change their order sizes. The hope is that with AI, Kraft Heinz workers can recognize patterns and precisely address these demand fluctuations, boosting sales.

For example, demand for pantry items increased sharply during the pandemic as consumers stayed home and stocked up. For years, food manufacturers had a strong enough position to hike prices even beyond what was needed to absorb inflationary costs, boosting their bottom lines. More recently, manufacturers' price-hiking power has decreased.

"Value is top of mind right now for the consumer," Carlos Abrams-Rivera, Kraft Heinz's CEO, told investors during the company's second-quarter conference call in July.

Kraft Heinz's net sales declined year over year in the first two quarters of 2024. Revenue has softened at many large food companies, including WK Kellogg, General Mills, and Mondelez, as volumes drop as consumers have been pushing back on nearly three years of steady price increases.

Big food companies want to use AI to make their supply chains efficient

The pandemic caused massive supply chain disruptions for food manufacturers, ranging from shutdowns at meat processing plants to sharp jumps in demand for self-stable products, resulting in higher competition for raw ingredients and packaging materials. Following the disarray, Kraft Heinz and its rivals began investing more in their supply-chain technology.

Several large packaged-food companies, including Coca-Cola and Conagra, have in recent years announced supply-chain projects — many of which involve AI — aimed at boosting efficiency.

Last year Kraft Heinz said it added $30 million to sales by applying AI and the Lighthouse to its supply-chain functions, including using technology to preemptively flag when service may be disrupted. It hopes to capitalize on retailer-demand-forecasting technology next.

Davis said Kraft Heinz was using data from big-box retailers fulfilling online orders to give food companies more frequent updates on their packaged-food-item availability.

Working with Microsoft to make factories more streamlined

Kraft Heinz has sought to stay dominant in food manufacturing partly by acquiring numerous brands over the years.

Thanks in part to its history of acquiring other brands, many of Kraft Heinz's factories have used different industrial computers, assembly lines, and machines. Over a few years, it pulled data from its various manufacturing systems, determined which data would be most useful for supply-chain and demand-forecasting recommendations, and put it in a centralized system, Lighthouse, which it launched in 2022 and uses in North America and Europe.

Lighthouse pulls in data from software including Intenseye, a workplace safety tool, and ST-One, which tracks inputs from sensors attached to manufacturing systems in North America, Europe, and Brazil.

Kraft Heinz is using Microsoft Azure's cloud platform as it pulls together all data and analytics into one home. The company also worked with Microsoft for a new generative-AI tool called KraftGPT, which employees are using as a catch-all "Ask Me Anything" type of app.

Creating a more autonomous workflow

Gustavo Gotelip, the vice president and head of supply-chain operations for Kraft Heinz's food-service and ingredients divisions, told BI that the company's investments in data collection had led it to retrain manufacturing teams.

Gotelip added that the company was hiring and training more data and automation engineers and robotics experts, though he declined to expand further.

Kraft Heinz says its investments in automation and AI aren't meant to replace its workforce. A representative said the company viewed technology investments as a "copilot," using technology to make more informed business decisions across manufacturing, supply chain, and logistics.

Davis said it had been easier to onboard talent because Lighthouse gives employees directions for handling demand disruptions.

"It's almost like you can take a person from day one and make them just as good as a person that's been there 10 years," Davis said. "Because the system is telling you exactly what you need to do."