Tyson Foods Improves Customer Experience With AI-Powered Innovation

Consumers experience Tyson Foods through a broad portfolio of iconic products and brands that include Tyson®, Jimmy Dean®, Hillshire Farm® and Ball Park®. But what the customers, consumers, and foodservice operators we serve can’t see is the innovation that extends far beyond our test kitchens. As part of a multi-year digital evolution to connect more closely with customers and consumers while growing our brands and operating efficiently, the Tyson Foods Global Technology team is radically improving how customers discover our products and recipes by leveraging the power of generative AI on Amazon Web Services (AWS).

Generating richer results through semantic search 

The Tyson Foods digital evolution began with a crucial first step: improving how foodservice operators explore an estimated 600 recipes and 1,200 active products on tysonfoodservice.com. Online search is especially critical for smaller foodservice operators with razor-thin margins who are looking for new ways to optimize their menu or achieve back-of-house efficiencies through purchase decisions like buying cooked chicken breast instead of raw.

The existing search functionality was a basic index keyword search that relied heavily on typing in exactly the correct phrase. The need was clear — provide more robust access to information, but also contextual product guidance and tailored recommendations. Improving online search was the catalyst for embracing generative AI, setting the stage for a more intuitive browsing experience that matches online behaviors of younger audiences, an increasingly important demographic for Tyson Foods. 

Building a scalable generative AI framework

As an early adopter of generative AI technology, the Tyson Foods Global Technology team developed a solid foundation to leverage AI for employees and customers. This foundation enabled the AWS Generative AI Innovation Center and the Tyson Foods technology team to partner on developing and deploying the new AI-backed search experience for tysonfoodservice.com in less than three months.  

Tyson Foods leveraged Amazon Bedrock, a fully managed service from AWS, to further scale the framework by adding access to a variety of open-source Large Language Models (LLMs) through a single, secure API.  

Tyson Foods selected Amazon Bedrock for multiple reasons:

• It offers access to multiple high-performing foundation models (FMs) from leading AI companies through a single API and a broad set of capabilities to build generative AI applications. This makes it easy to experiment and evaluate various LLMs to ensure the best performance and cost for each use case.
• Amazon Bedrock can be used to customize models for specific tasks using our company's own labeled datasets, empowering Tyson to deliver a personalized user experience.
• Amazon Bedrock has security, privacy, and responsible AI capabilities built in. 

The semantic search solution uses Amazon Titan Embeddings to power more accurate and contextually relevant multimodal search, recommendation, and personalization experiences. It also uses Anthropic Claude Sonnet for natural language processing and Amazon OpenSearch Serverless for scalable search capabilities. 

Better results, greater customer insights

The new system is dramatically improving the relevance of search results by transforming the way foodservice operators interact with the Tyson Foods digital platform. 

This enhancement goes beyond influencing purchases; it offers Tyson Foods invaluable insights into the needs of our foodservice customers. By analyzing search patterns and interactions, we will be able to better understand and anticipate customer demands. 

The partnership with AWS Generative AI Innovation Center accelerated time-to-value of a key generative AI-based project and helped our team unlock additional capabilities. The center’s focus on scalability and production readiness streamlined integration into our enterprise production system and provided a pattern to easily incorporate additional generative AI functionality into other applications.

Best practices to ease implementation

The speed and seamlessness of this project suggests a few ways to help ensure success:

• Make sure your IT department’s digital enablement organization partners closely with business unit stakeholders for projects so closely related to customer experience.
• Reach out to strategic technology partners, as we did with AWS, to explore how they can help you test and learn new concepts quickly.
• Evaluate AI as you would any other technology — build a case for how the business will use the functionality rather than applying it broadly. This project was the winner of an internal use case “bake-off.”

What’s next for Tyson Foods

Improving the foodservice operators search experience is just the start. The Global Technology Team will leverage our generative AI framework to continue bringing to life our customer and consumer obsession by building personalized, nurtured journeys to move consumers through the purchase funnel. Our immediate priorities are developing chatbots for natural conversations and adding semantic search to our consumer-facing brand sites.

This project illustrates Tyson Foods’ position as an innovation leader in a very low-margin industry with historically low rates of technology investment. AI is helping us create stronger connections with customers and grow our business while maintaining our margin structure. As we continue to push the boundaries of innovation, it brings us one step closer to our mission: bringing high-quality, affordable food to every table in the world.