What interests me about personal AI agents is what happens on the other side of the shopping conversation. An agent may know the shopper’s needs and preferences. How does the brand help it understand which product fits, and why?
We are seeing tools such as Meta Muse, OpenAI Dots and Instinct emerging. We’re getting a glimpse of what that journey could look like. An agent can carry a shopper’s needs and preferences into the search for a product, explore the options and act on their behalf. For brands, that means another participant in the shopping conversation.
That question is at the heart of the latest DaVinci Commerce announcement. We’re expanding our Agentic Commerce Experience Platform with Brand Agent and our patent-pending Product Context Memory (PCM) technology.
Rajat Agarwal, Senior Managing Director of Accenture Song, raised a closely related question in his recent LinkedIn post: how does a brand get chosen when shopping starts with a personal agent? He points out that an agent needs to be able to find a product, understand it and ultimately buy it.
One line stayed with me: “Brands may need to win twice: first with the [consumer], then with their agent.” That captures why I see product knowledge and the brand experience as connected parts of this new shopping experience.
It also raises a practical question: how should brands think about “marketing” to agents? How should brands prepare for their new role in the shopping journey? Accenture’s research, “Agentic Commerce: Make Your Brand Unmissable,” offers a useful way to think about this. It describes horizontal agents like ChatGPT, Gemini and Claude, that work across categories, and vertical agents that go deeper within specific domains such as beauty, travel or financial services.
For brands, a practical place to start is the product knowledge these horizontal agents, AND personal AI agents like Muse, Dots, and Instinct rely on. Will an agent understand how a product fits the shopper’s situation well enough to recommend it?
Brand Agent gives brands a way to participate in that exchange
Think of DaVinci Commerce Brand Agent as the brand’s representative in a conversation with a consumer’s AI agent. It helps the agent understand the products using trusted information from the brand, shared in formats it can work with, including ACP, UCP, JSON-LD and MCP.
Behind Brand Agent is Product Context Memory, a persistent knowledge graph for every SKU. It connects what we know about a product with why someone might want it: what they need it for, which benefits matter to them and their personal preferences.
Someone asking for a jacket for a rainy October trip is describing a situation. The product information needs to help an agent connect that situation with relevant features, such as waterproofing (or only water resistant); warmth and material’s breathability for hiking or other more intense activities; comfort and packability.
Our CEO, Diaz Nesamoney, put it this way:
“Consumers describe their lives, their needs and the problems they want to solve to personal AI agents. Most product catalogs still speak in technical specifications. Product Context Memory connects those two worlds. Our Agentic Commerce Experience Platform gives brands and retailers the infrastructure to continually build that context, make it available to AI agents and measure whether it is actually helping their products get discovered and chosen. As AI agents become a new gateway to commerce, the quality and depth of product knowledge will increasingly determine which products earn a recommendation.”
From product knowledge to purchase
Together, the capabilities of our platform support the journey from discovery to purchase:
Discover: Agentic Discoverability Engine enriches product catalogs with consumer context and measures SKU-level discoverability.
Experience, then choose: Agentic BrandStore enables curated, conversational brand and retail experiences for consumers inside AI platforms, giving consumers a reason to choose the product.
Buy: The platform provides a seamless handoff to multiple purchase paths, including a brand’s online store, retailer websites, the nearest physical stores or in-app checkout systems, making it a closed-loop path from conversation to purchase.
To me, this is the emerging model of agentic commerce: a shopper’s agent translating intent and a brand’s agent responding with the right product knowledge and experience.
It is the beginning of an agent-to-agent path from conversation to commerce. Brands and retailers must now speak not just to consumers, but to AI agents too.
The questions we’re exploring on the podcast
These are also some of the many questions we’ll explore in my new AI Commerce Innovators podcast. The podcast, launched on October 7, covers the broader challenges and opportunities brands and retailers need to address as agentic commerce evolves.
In the inaugural episode, I spoke with Chip Knicker of Accenture, Diaz Nesamoney of DaVinci Commerce and Emily Pfeiffer of Forrester about how shopping is becoming a conversation with AI. We explored what that means for how shoppers discover and compare products, and how brands help AI understand their products while creating an experience that feels like their brand to move shoppers towards purchase.
Chip offered a useful way to think about the choices ahead. Every brand must first become the choice of agents by showing up in the horizontal systems where consumers are asking questions. Some brands may go further and become an agent of choice by creating a specialized, brand-controlled experience. He pointed to Accenture’s work with DaVinci Commerce and clients on Agentic BrandStore as an early example of brands testing and learning their way into that second path.
Together, Chip, Diaz and Emily explore what this means for discoverability, the new agentic shelf and the parts of the shopping journey moving outside brands’ traditional environments.
Learn more about DaVinci Commerce announcement: https://www.davincicommerce.ai/press/davinci-commerce-brand-agent-product-context-memory/
Check out Rajat’s Weekend Musings: The personal shopping agent is here. https://www.linkedin.com/feed/update/urn:li:activity:7512481741346041856/
I also recommend Accenture’s “Agentic commerce: Make your brand unmissable,” written by Accenture Song CEO Ndidi Oteh, Rajat Agarwal and the Song team. https://www.accenture.com/en/insights/song/agentic-commerce
Our first episode “Nobody’s browsing anymore. They’re asking.” arrived on October 7. Listen to Chip, Diaz and Emily discuss what agentic commerce means for brands and retailers. The podcast is made possible by DaVinci Commerce in partnership with AI Realized. https://davincicommerce.ai/video/nobodys-browsing-anymore-theyre-asking/
