Webinar Recap: Beyond the Control Tower — Orchestrating Food and Beverage Supply Networks with Agentic Decision Intelligence

Summary

In our Future.Now webinar, “Beyond the Control Tower: Orchestrating Food and Beverage Supply Networks with Agentic Decision Intelligence,” we explored how food and beverage supply chains are evolving to meet a faster, more fragmented, and more volatile market. Consumer preferences are shifting rapidly, while cost pressures from ingredients, tariffs, and logistics continue to rise. These forces are making traditional approaches, centered on visibility alone, increasingly insufficient.

We examined how leading organizations are moving beyond control tower models to build decision-centric supply networks. Rather than simply monitoring data, these networks continuously sense changes, evaluate options, and take action across inventory, suppliers, and fulfillment. This shift enables faster, more coordinated decision-making at scale, helping organizations respond in real time and protect margins in an increasingly dynamic environment.

Key Takeaways

• Visibility alone is no longer enough to compete.
  Traditional control towers provide insight into what is happening, but they stop short of enabling action. As supply chains grow more complex, the real differentiator is the ability to act on that information quickly and consistently. Organizations that translate visibility into decisions are better positioned to respond to disruption and maintain performance.
• Decision velocity is becoming the new performance driver.
  In food and beverage, delays quickly translate into lost sales, excess inventory, or margin erosion. Faster decision-making allows organizations to adjust inventory positions, reallocate supply, and respond to demand changes before issues escalate. Speed, paired with accuracy, is now central to operational success.
• Supply chains are shifting from linear processes to decision networks.
  Instead of isolated decisions made within silos, leading companies are connecting decisions across the end-to-end supply network. This allows inventory, sourcing, and fulfillment decisions to be coordinated in real time, reducing friction and improving outcomes across the entire value chain.
• AI-driven orchestration enables coordinated action at scale.
  Advanced systems are now able to evaluate multiple variables simultaneously and recommend or execute decisions across functions. This orchestration ensures that actions taken in one area, such as inventory deployment, align with broader supply and demand conditions, creating more consistent and effective outcomes.
• Early ROI builds momentum for transformation.
  Organizations are seeing measurable impact within the first 90 days by focusing on high-value use cases. Early wins help demonstrate value, build confidence, and create a self-funding model for scaling decision intelligence across the enterprise.

Speakers

Ram Krishnan, SVP, Platform Product Marketing, Aera Technology Ram leads the GTM strategy for Aera Decision Cloud with over two decades of experience in enterprise, software, product strategy and customer success. Ram brings a deep understanding of how to align solutions with customer values. Under his leadership, Aera has earned a strong reputation for delivering measurable business outcomes and customer satisfaction.

Sean Harapko, Consumer Products Growth Markets & Beverage Sector Leader, EY Sean leads EY’s Americas consumer products growth and beverage sector. He brings more than 30 years of experience driving transformation across the industry. Sean works with organizations to accelerate growth, optimize supply chains, and lead large-scale operational and digital change initiatives.

Full Recording

Access the full webinar recording here.

Q&A

At the end of the presentation, we held a short session dedicated to answering attendees’ questions. Below are several of the key questions and answers.

Q: For a business with thousands of SKUs and a very complex global supply chain, how practical is it for an end user to analyze hundreds or thousands of Aera recommendations daily and either modify or accept them?

A: That’s a great question, and it gets to the core of how the system is designed to work. The goal is not to have users manually review thousands of recommendations one by one. Instead, the approach is built around support, augmentation, and automation working together.

When the system surfaces a large number of opportunities, it becomes a question of where to apply automation. The system identifies those opportunities, and then business-defined rules determine which ones can be automated. For example, a simple rule might be: if the revenue impact is below a certain threshold, automate it.

These rules can become much more sophisticated. Organizations often define rulebooks by region, category, or business priority, and those rules are applied dynamically within decision flows. In one case, a deployment generated 10,000 recommendations in the first month. That revealed a long tail of smaller opportunities that had previously gone unnoticed. The natural next step was to automate a large portion of them using defined rules.

There are also prioritization techniques, such as applying Pareto principles or focusing on specific segments like key regions or high-value categories. On top of that, filtering tools allow users to focus only on the most impactful recommendations, such as those tied to the highest revenue.

So in practice, users don’t need to process everything manually. The system helps narrow the focus to what matters most, while automation handles the rest.

Q: When will agents take a fully automated route? Is that when confidence reaches 100%?

A: Not at all. In reality, a 100% confidence level doesn’t exist in AI, so waiting for that would mean waiting forever. Full automation doesn’t depend on perfection. It depends on thresholds that the business defines.

Each organization sets its own confidence levels based on the type of decision, the level of risk, and the potential financial impact. For example, a routine replenishment decision might be automated at 85% confidence, while a high-value supplier commitment could require 95% or even a human review.

The key point is that confidence is a trust indicator, not a measure of certainty. The system continuously learns from every accepted, modified, or rejected decision, which helps improve confidence over time. As that trust grows, automation naturally expands within the boundaries set by the business.

You decide what level of confidence is “good enough” for a specific type of decision to be executed autonomously. The system then applies guardrails within those parameters and takes action accordingly. Over time, as outcomes are learned and validated, those confidence levels improve, and the scope of automation can expand.

Q: Where does the data live?

A: Your data stays exactly where it is. It never leaves your environment.

The Aera agent operates within your firewall and connects directly to your existing systems, whether that’s SAP, ERP platforms, or data lakes. It doesn’t replace or move those systems. Instead, it connects to them through the Decision Data Model™, which harmonizes data in real time.

From there, decisions can be written back into those systems, effectively closing the loop between insight and action. So while the data remains distributed across your environment, the system creates a unified, decision-ready layer on top of it.

Q: How much effort is required for change management to drive adoption within supply chain teams?

A: From a practitioner’s perspective, adoption tends to be strong, often stronger than expected. When teams are faced with thousands of possible actions or optimization opportunities, there is a real appetite for tools that help prioritize and act more effectively.

Of course, change management still needs to address a common concern: if parts of the job are automated, what does that mean for the role? What is being seen in practice is not a reduction in the need for supply chain professionals, but a shift in focus. Instead of spending time on firefighting, teams are able to concentrate on higher-value decisions, such as innovation, strategy, and market positioning.

This shift often leads to role evolution. For example, demand or supply planners may move into broader, value stream–oriented roles. At the same time, many organizations still rely heavily on spreadsheets for day-to-day work. Moving to a system that not only provides visibility but also enables action creates clear value for users.

As a result, adoption is typically high, because the system helps people do more meaningful work and drive greater impact in their roles.