The arrival of generative AI is giving rise to a new sales channel: agentic commerce. It's no longer just an interface, but an intelligent agent capable of recommending and executing purchases. For businesses, positioning themselves on this channel has become a strategic imperative.
From multichannel to agent-based commerce: a major shift
The structuring of e -commerce has followed several major stages. Companies initially treated it as an independent sales channel before structuring a multichannel offering (merging their physical and digital sales channels). The introduction of artificial intelligence has led to the emergence of an additional sales channel: agentic commerce. Intelligent agents make purchase recommendations to consumers and then handle ordering the products they have selected.
The arrival of agentic AI in the e-commerce sector represents a significant disruption for businesses. Indeed, consumers using this channel no longer consult the brand's or seller's website, but rely on recommendations made by a chatbot, which automatically suggests a multi-vendor shopping cart containing the best products on the market. In some cases, these agents will even be able to make purchases entirely autonomously (for example, for recurring purchases).
To integrate this new sales channel, brands, distributors, and retailers must create bridges between their e-commerce platform and consumer chatbots (ChatGPT, Gemini, Le Chat, etc.). Smile supports companies in creating agents and masters key agentic commerce protocols, such as the MCP (Model Context Protocol) and the UCP (Universal Commerce Protocol). We also help optimize brand visibility with AI (GEO) and improve the user experience with UX-Republic , through a UXO (User Experience Optimization) approach.
B2B: Towards total automation of sourcing
AI is also expected to quickly find its place in B2B commerce. We're already familiar with "punch-out" catalogs, which allow purchasing teams to directly access their suppliers' catalogs from their e-procurement solution. As with B2C sales, a chatbot will enable these purchases in a multi-supplier environment. It will rely on specialized agents: one to check stock levels at each supplier, one to generate price quotes, and so on.
Supplier sourcing will also be accelerated. An agent will check the sales conditions of a potential supplier and request additional information if necessary. Ultimately, the entire purchasing process should be covered—and automated—by AI. The sales process will also be automated, with suppliers deploying agents capable of responding to the most common customer inquiries (sales conditions, stock availability, pricing, etc.). Furthermore, AI will allow for more precise sales forecasting, enabling companies to anticipate demand peaks.
Smile: your trusted partner for open source AI
The Smile teams are experienced in working on data structuring. This is essential for agent-based commerce, which requires real-time and reliable information on elements such as product codes, prices, and inventory. For all these reasons, we believe that the future of technical architectures must be modular and composable, and when possible, open-source solutions should be chosen (e.g., Medusa ).
Another key theme is data privacy and the sovereignty of the solutions implemented. At Smile, we ensure that only the necessary information is communicated to chatbots, so you can maintain maximum control over your data. Regarding sovereignty, we rely on European or French infrastructures to deploy your solutions.
Finally, let us remember that our particular positioning around digital autonomy and our transparency in developments make us more than just a partner: Smile aims to be the trusted third party for your IT projects!
