A panel of AI experts explored how artificial intelligence is transforming pet food operations during "Step Into the Future: The AI Journey from Recipe Design to Kibble Production" at the American Feed Industry Association's Pet Food Conference, held during the International Production & Processing Expo (IPPE) 2026 in Atlanta, Georgia, U.S.
Host Eric Altom, Ph.D., director of innovation for companion animals and technical nutritionist at Balchem Animal Nutrition and Health, moderated the discussion with Hana Bieliauskas, senior vice president and partner at Inspire PR Group; Johanna Ballesteros, animal science and AI thought leader at SWARM Engineering; Filip Snauwaert, solutions architect at BESTMIX Software; and Tara Zedayko, chief scientific officer at Ollie Pets.
Here are 10 takeaways from the AI panel discussion:
1. Data readiness shouldn't delay AI adoption
One common misconception is that companies aren't ready to implement AI because their data isn't perfect. "From experience approaching customers working on use cases, one common denominator is 'we are not ready with the data. The data is not ready for AI,'" said Ballesteros. "But usually, if you delay that, if you think that data is not ready, then you delay value as well."
2. AI amplifies expertise rather than replacing jobs
The fear that AI will eliminate human roles is misplaced, noted the panel. Today's AI applications are designed to scale scarce expertise and make current employees more efficient. Zedayko explained that at Ollie, AI helps scale expertise that's extremely limited: "If you think about veterinary nutritionists, for example, there are less than 100 in the country. If there were more, maybe AI wouldn't be as useful."
3. Extrusion process optimization delivers measurable results
Real-world implementation of AI in pet food extrusion has shown significant operational improvements. Snauwaert shared results from a proof-of-concept project: "We immediately saw that when a new product is being produced, the extrusion process starts. They can start up a lot faster." The company achieved up to 33% reduction in rework and 50% less moisture swing.
4. Trust and accuracy are critical barriers
For pet food manufacturers, AI predictions must be completely reliable since errors could affect product safety and quality. "Pet food producers cannot afford even one error, because then, all of a sudden, the pet food they make cannot be produced, or it's not healthy for the animal," Snauwaert explained. Data security and maintaining proprietary recipe information also create significant trust barriers.
5. Start with clear, focused use cases
Successful AI implementation begins with identifying specific problems where AI can deliver measurable return on investment. The experts recommend focusing on processes that require significant human expertise or where efficiency could be dramatically improved, such as production planning, quality assurance, or formulation optimization.
6. System integration solves data silo challenges
Many pet food companies struggle with data trapped in separate systems. "It's very important that those systems or solutions integrate with existing data," Ballesteros noted. Connecting formulation software, production systems, and quality control platforms allows AI to optimize across the entire operation rather than in isolated steps.
7. Domain-specific AI ensures safety and compliance
For regulatory and safety-critical applications, AI systems must be limited to verified, domain-specific information. The American Association of Feed Control Officials (AAFCO) uses an AI assistant called Ava that only accesses AAFCO source materials, ensuring accurate regulatory guidance while maintaining proper attribution and preventing misinformation.
8. Simple applications offer quick entry points
Companies can begin experimenting with AI through low-risk applications like summarizing reports, generating content ideas, analyzing consumer sentiment from reviews, or identifying patterns in product feedback. These applications help teams become familiar with AI capabilities before tackling more complex manufacturing or formulation challenges.
9. Marketing applications require ongoing training
The pet food industry is still developing expertise in AI-powered marketing tools. "A lot of people are not trained on them effectively yet," said Bieliauskas. "That is changing because AI has become a huge priority in communications."
10. Photo-based health screening enables personalized nutrition
AI-powered image analysis is making personalized pet nutrition more accessible. Zedayko described how Ollie uses photo submissions from pet parents: "Members can take a photo of their dog. It can be their stool, their skin and coat, their body condition, or their teeth. Then through AI and/or expert assistance, we will provide personalized health plans." The company has processed over 100,000 photos from more than 54,000 dogs.
