Challenge

Seed Digital is a technology company specializing in intelligence for physical retail. Its platform combines 3D sensors and AI to measure customer traffic in stores, analyze purchasing behavior, and deliver operational intelligence to retail chains. All in real time, without storing biometric data.

The model works well for what's seen on the sales floor, but there's another side to the operation that doesn't appear on the sensors: the performance of each salesperson. Individual conversion rate, items per transaction, average ticket; indicators that depend on structured data collection and a consistent monitoring process.

The feedback process in retail has a characteristic that makes it especially costly: it doesn't happen once. It happens every week. For each salesperson on the team, the manager needs to review the period's indicators, interpret what the numbers say about that specific professional, and translate that analysis into a concrete action plan. Done rigorously, the process demands analytical attention and communication skills simultaneously.

In retail chains with dozens of stores, inconsistency among managers is inevitable: some follow the process rigorously, others superficially, and the result for the company is performance management that varies more by manager than by store.

Seed Digital recognized this pattern and brought in Evo Systems to solve it.

Solution_

The answer wasn't a layer of reports on top of what already existed. It was a product built from scratch to close the entire cycle: from the indicator to the action plan.

Evo Systems developed Alia.X as an app and web system integrated into the Seed Digital platform, focusing on two objectives: presenting performance indicators clearly to the manager and automating the generation of feedback based on this data.

The core functionality of Alia.X is the generation of feedback through artificial intelligence. The manager accesses the salesperson's profile, views the performance indicators for the period, and receives AI-generated feedback based on real data. The action plan emerges from the data, not from the subjective perception of the evaluator.

Building the business logic was half the job; the other half was designing an experience that would work in the real-world environment of a physical store, where the pace is different from an office and the salesperson can't stop to browse menus.

In this way, Evo Systems understood how the sales floor truly works, thus improving the salespeople's experience. Alia.X was designed to digitize this practice without altering the team's behavior.

The salesperson arrives, walks to the tablet installed in the store, and starts their shift with a personal password. From there, their name is added to the list. When it's their turn, they update their status to "Serving a Customer" directly in the app, where the manager can monitor it in real time.

The customer service closing workflow was designed with the same care. Instead of an open field, the salesperson records the result in a structured way: did the customer buy, not buy, was it an exchange? If not, what was the reason? This structure transforms each interaction into usable data. The navigation was designed in two distinct layers — a streamlined interface for the salesperson, a consolidated view for the manager — with the right information for each profile, without visual clutter between the two.

Tools used_

• Flutter (Dart)
• .NET
• AWS 
• AI.

Flutter was chosen for developing the screens in both the mobile app and the web system—the same codebase for both environments, ensuring visual consistency and less maintenance effort. The APIs were built in .NET and hosted on AWS with serverless functions, an architecture suitable for the usage pattern of a retail network: predictable access peaks at the beginning of each week, without the fixed cost of idle infrastructure for the rest of the period.

Testimonial_

"We have been working with Evo Systems for 3 years, during which time we have developed digital products with their support. The partnership has been essential to accelerate the development and quality of deliverables, especially thanks to the support of a UI/UX team, which ensured the creation of solutions aligned with our expectations."

- Sidnei Raulino, Seed Digital

Do you still do all of this manually?

Whether in retail or any operation with teams and goals, evaluation processes often consume more time than they should and not always with the same quality standard.

Today it is possible to transform data into actions in a faster, more consistent and practical way.

Want to understand how this can work in your context? Talk to Evo Systems.