Agentic AI and the Not-So-Quiet Retail Revolution

For decades, retail software followed a simple rule: it did exactly what we told it to do. Nothing more, nothing less. Hard-coded logic governed everything from inventory allocation to checkout flows. But as retail has become more complex and more unpredictable, that model is now being challenged.

Today, we are entering a new era of agentic AI. Unlike traditional systems, agentic AI doesn’t just execute instructions, it can reason, plan, and act in the gray areas where static rules fail. In short, retail is moving from programmed to autonomous.

That shift was at the heart of a Big Ideas session at NRF 2026, featuring Sanjeev Siotia, CTO of Manhattan, alongside retail and commerce leaders Karen Beebe (CTO, Bealls, Inc.), Dave Stevens (CTO, Groupe Dynamite), and Eduardo Frias (Field CTO, Shopify). Together, they explored what agentic AI really means for retailers.

Seriously, What Is Agentic AI?

At its core, agentic AI is about outcomes. Instead of writing detailed logic for every scenario, retailers can define an objective and allow AI agents to determine how to achieve it. These systems operate in what Siotia described as a “think–see–do” loop. “They assess the situation, make decisions, and take action.”

Retail can be messy, so the appeal is obvious. Demand shifts unexpectedly, weather disrupts supply chains and consumer preferences change overnight. Not surprisingly, hard-coded, rigid systems struggle with exceptions, whereas agentic AI is designed to handle them.

From Task Automation to Outcome Ownership

Interestingly, when asked what task they would immediately hand off to an AI agent, panelists didn’t focus on flashy use cases. They talked about emails, scheduling, presentations - work that consumes time but adds limited strategic value.

This theme carried into the broader discussion - agentic AI isn’t just about speed, it’s about freeing human beings to focus on what truly matters.

For Groupe Dynamite, speed and complexity are inseparable. According to Dave Stevens, the real value of AI lies in tackling problems that are simply too complex for humans to solve quickly, such as demand planning, personalization, and real-time decision-making across channels.

What was notable however, was where leadership focus is directed. Rather than starting with cost savings or labor automation, Groupe Dynamite is prioritizing AI initiatives that improve customer experience and drive revenue. Automation follows naturally and as Stevens put it, “when you build for growth, efficiency comes out of the box.”

AI is a Cultural Shift, not Just a Tech Project

Bealls, Inc., a 111-year-old, family-owned retailer, offered a different but equally instructive perspective. Karen Beebe emphasized that AI adoption can’t be technology-led alone. It must be embraced across the organization.

At Bealls, the focus is on using AI (and increasingly agentic AI), to support decision-making during critical moments, like the ‘Monday Morning Planning Rush’ that defines retail operations. Agents can ingest data, analyze plans, surface recommendations, while keeping humans in the loop. Over time, this changes not just processes, but fundamentally how teams work too.

Both Karen and Dave underscored the same truth: change management is harder than the technology itself. Many digital transformations fail not because the tools don’t work, but because people don’t trust or adopt them. Successful AI strategies start with business engagement, education, and a willingness to experiment.

The Democratization of Capability

From Shopify’s vantage point, agentic AI represents a great equalizer. Eduardo Frias described AI as a “multiplier of human ambition,” one that gives small merchants access to capabilities once reserved for enterprises with massive IT teams and budgets.

Shopify’s AI assistant, Sidekick, is built directly into its platform, not bolted on – much like Manhattan’s Agent Workforce is native to the various solutions on the Manhattan Active® Platform. It uses the same underlying intelligence whether the merchant is a global brand or a solo entrepreneur. From generating product descriptions to building custom apps through natural language, AI removes technical barriers and puts power directly into the hands of business users.

This democratization has major implications for the SaaS landscape. As Dave noted, when English becomes the new programming language, the ability to build, test, and deploy solutions accelerates dramatically. Companies that embrace this shift will move faster. Those that don’t, may struggle to keep up.

Customer Experience in an AI-First World

Agentic AI is also reshaping how customers discover and buy products. Traditional browsing-based e-commerce is already giving way to conversational, intent-driven experiences. Increasingly, discovery happens through AI search rather than traditional search engines, while authenticity, community, and user-generated content matter more than ever.

Still, the panel cautioned against optimizing purely for efficiency. Shopping isn’t always about completing a task. There is joy, exploration, and social connection involved. The challenge and opportunity for retailers is to use AI to enhance that experience, not eliminate it.

Guardrails Without Killing Innovation

Of course, giving software more autonomy introduces risk. An AI agent that hallucinates isn’t just embarrassing it can be costly. Governance, data accuracy, and human oversight are all essential requirements.

There was broad consensus that trust starts with data. Stevens commented: “Get your data straight first. If it’s garbage in, you get garbage out, no matter how good your AI is.”

The Journey Starts Now

If there was one unanimous message on show during the Big Ideas session, it was around urgency. Retailers cannot afford to sit this out. Even small pilots matter. Learning how to prompt, experiment with agents, partner with vendors, and share insights with peers is central to building momentum.

As Siotia concluded, the formula is simple, but not easy: “start now, get your data right, and learn together.”

Agentic AI isn’t a sci-fi concept any longer, it’s here today and it’s already reshaping retail. In a brave new world, framed by innovation and speed, those organizations that engage early will be the ones who define what comes next.