Retail
Black Friday: Are Your Systems Ready?
The discussions at SXSW 2026 confirmed a trend we’ve already observed in most projects: AI has moved beyond supporting the user interface and is now at the heart of decision-making.
The concept of Agentic Commerce has evolved. Whereas we used to talk about bots that answered questions and suggested products, today we’re talking about systems that already make critical decisions—from inventory to pricing—without relying on direct human intervention at every step.
In practical terms, this means that today, an agent is already able to automatically reorder recurring items—for example, by choosing between suppliers based on price, delivery time, and availability—without any user interaction.
The impact of this change is both financial and operational.
Traditional digital retail was built on the premise of human interaction with screens. But when customers become AI agents that make scheduled purchases based on data and user preferences, the experience shifts from the screen to taking place between systems.
In this scenario, efficiency is measured by the infrastructure’s ability to respond in real time to thousands of simultaneous decisions that impact revenue. This ranges from dynamically adjusting prices to prioritizing high-margin orders—all within seconds, without manual intervention.
The problem is that many companies still try to support this new dynamic on platforms that were designed for a different market environment, such as legacy ERPs.
In practice, this results in delays, inconsistencies, and eroded profit margins—not due to a lack of strategy, but because of structural limitations. Take, for example, a stockout that could have been prevented but isn’t, because the system cannot update inventory in real time, while a competitor is already automatically adjusting supply and distribution.
The key difference between leaders and followers in 2026 is the use of AI to solve problems from the inside out.
Instead of simply piling new solutions on top of old structures and systems, market leaders use generative AI to document, refactor, and modernize the code of legacy systems. This approach transforms technical debt into real scalability.
In practice, this reduces the time required to analyze legacy systems from weeks to hours, speeding up decisions that previously hindered the development of digital products.
By applying AI to the ERP foundation, it is possible to create layers of integration that enable agent-based intelligence to access inventory, logistics, and pricing information in real time, with the security and governance that the corporate environment requires. This enables automated decision-making based on reliable data without disrupting operations.
This allows a system, for example, to avoid selling a product that is at risk of logistical delays by automatically redirecting the order to another region or supplier.
Technical stability and system modernization are no longer the exclusive domain of IT; they have become topics of discussion in the boardroom.
In today’s retail landscape, a system’s inability to process a user-driven decision literally means a lost sale to a competitor with a more resilient and integrated infrastructure, as well as the company’s exclusion from recommendation systems. In a scenario of automated decisions, failing to respond in a timely manner is no longer merely a technical glitch but translates directly into lost revenue.
The question for executives today is no longer “which AI should I use?”, but rather “does my infrastructure support the level of autonomy the market demands?”.
At Inmetrics, we understand that the success you see on the front end—whether in a seamless transaction or an efficient buying agent—stems from the technical rigor applied at the core of the operation. That’s why we help companies transform their legacy systems into platforms ready for the agent-based economy, connecting business strategy with technical execution at the code level.
See how Change Makers are reshaping performance and efficiency in the retail sector.
If you enjoyed this analysis, please share it with your network and keep following our insights and perspectives on the evolution of engineering and AI.
