Step Into the Future - AI with Johanna Ballesteros, SWARM Engineering and Guests

The episode opens with an introduction to the panel and a high-level overview of AI’s rapid acceleration across food and feed systems, setting the stage for a discussion on separating real-world applications from hype. (00:00 – 00:50)  The panelists then outline current AI trends, noting that while adoption in the pet food industry has lagged slightly, companies are increasingly leveraging AI for customer insights, marketing optimization, and early-stage operational improvements. (00:50 – 04:00)  The conversation shifts into one of the most critical challenges: trust in AI systems, including issues like hallucinations, accuracy, and the need for domain-specific models to improve reliability and decision-making confidence. (06:22 – 10:09)  From there, the discussion explores how AI is already delivering value in manufacturing and plant operations, particularly through tools that capture institutional knowledge, reduce downtime, and support faster root cause analysis. (10:09 – 13:58)  The panel also highlights the role of AI in decision intelligence and supply chain optimization, where advanced models help companies respond to volatility, improve forecasting, and make faster, data-driven operational decisions. The conversation then turns to the consumer experience, explaining how AI quietly powers personalization—from e-commerce recommendations to pet health insights—often without users even realizing it. (16:36 – 19:59)  The panel addresses concerns around job displacement and data security, emphasizing that AI is best viewed as a tool that augments human capability, while also acknowledging the growing importance of protecting intellectual property and proprietary data. (19:59 – 27:24)  Looking ahead, the discussion explores the emergence of agentic AI systems, where autonomous agents can not only generate insights but also take action—executing decisions, optimizing processes, and closing operational loops in real time. (30:44 – 36:41)  The episode concludes by reinforcing a key takeaway: companies don’t need to solve everything at once—the best way to begin is by identifying a meaningful problem and starting there, building capability, trust, and ROI over time. Please subscribe and share this episode with your industry colleagues and invite them to join us at the Real Science Exchange virtual pub table. Be sure to register for upcoming Real Science Lecture Series webinars to continue the conversation.  Finally, if you’d like a Real Science Exchange t‑shirt, screenshot your rating, review, or subscription and email it to [email protected] with your size and mailing address, and we’ll send one your way.