By: Anita Beveridge-Raffo
The food and beverage sector spent nearly $12 billion on AI in 2024 as brands seek automation and efficiency in navigating supply chain challenges, meeting safety requirements, and keeping up with the world’s growing demand for food. By 2030, annual spending on AI in the sector is projected to reach more than $85 billion, according to San Francisco-based Grand View Research. Clearly, brands have embraced the technology.
The big shift today is how brands are using AI to meet their goals. Ambitions are growing, and executives are wondering, “How can I grow my top line at the same time as driving efficiencies to the bottom line?” Increasingly, AI is more focused on automation, not just for physical processes like sorting and packing, but also when it comes to data, planning, and strategy. At the same time, brands are embracing AI systems that are more granular, taking into account every package or product, and are using AI systems to connect operations and strategy across departments.
Automation Starts with Humans in the Loop
More brands are discovering that AI systems need to be able to adjust themselves on the fly, to complete tasks rather than simply provide information in order to bring significant business value. For the last few years, many food manufacturers have used complex, highly tuned ML models to update and optimize their logistics plans. But they often tell me that by the time the process is completed, the plan is out of date. This is usually because humans are required to complete the tasks in between systems, like taking data from one system, reviewing it, and uploading it to another – a common symptom of the adoption of highly specific but difficult to integrate point solutions over many years.
Now, more companies are using a series of chained AI agents and APIs to complete these previously human-based tasks, truncating the time it takes to update and optimize processes like logistics plans, and therefore are able to adapt more quickly to changing realities on the ground. Unlike human chat-based interactions with large-language models (LLMs), agents do not need to be prompted. They act on their own, based on triggers from refreshing data and model outputs, and according to certain parameters and business rules. While humans still closely supervise such systems, people are moving away from serving as data pipelines or actively giving prompts, and are increasingly letting automation take over.
The Value is in the Details
More brands are using AI that thinks and acts not in terms of aggregated categories of products, but down to the SKU level. For example, Anheuser-Busch InBev SA/NV can see how single cans of Budweiser versus six-packs are selling to wholesalers in a specified geographic area. This is actually a reversion back to how brands used to track and think about their supply chains before many of them hit the scale they’re at today.
As highly scalable AI models unlock the ability for brands to take their analytics to a more granular level, this data and insight also allow for more effective marketing campaigns and promotions. A large packaged goods company, for example, is moving toward running promotions based on current market sales, rather than looking back at last year’s data for a certain season or geographic market and basing a campaign on that, a practice that seems to be the industry standard.
Improving Business Agility with Cross-Departmental Insights and Cooperation
As they deploy more AI, brands are realizing that it can eliminate some of the siloes that contribute to supply chain challenges. Beyond connecting sales data to promotional planning and budgeting, cross-departmental use of AI can also contribute to innovation in recipes, product design, and packaging. This is especially helpful as consumer preferences are changing more rapidly than ever. Rising prices, changing diets, whether for health reasons or due to the growth of GLP1 drugs, are disrupting shopping routines and requiring that brands understand exactly what people are buying – or not buying – and how these patterns can help predict, plan for, and get ahead of future trends.
The push by boards and executives to adopt AI is only growing. Going forward, those companies that shift their thinking to automation by AI agents, granular data, and using AI systems to integrate operations, strategy, and planning across departments are on track to become more agile, resilient, and efficient. However, implementing these types of unified AI approaches can rarely be done at the departmental level; instead, they require visionary leaders who have the vantage point of looking across the entire company. That approach, rather than just focusing on individual departments or challenges like supply chains, is how food and beverage brands are able to build real return on their AI investments.
Anita Beveridge-Raffo is a lead deployment strategist at Palantir Technologies. Anita joined Palantir in 2015 and is responsible for technical deployments of Palantir software across retail and consumer packaged goods.
