Selectio

Introduction

Selectio is a B2B toy trading company that has spent more than fifteen years connecting factories in Asia with large retail chains and importers, with commercial presence across nearly every country in South America. Its clients include Walmart, Falabella, Coppel, Lojas Americanas and Tottus, some of the largest retail and department store chains in Latin America. Its role goes well beyond the sale itself: Selectio supports each import from product development and testing through certification, quality control, logistics and after-sales follow-up.

A buyer at a large retail chain chooses among thousands of products from different factories and needs fast answers: what each item is, what it is made of, how much container space it takes and what it costs. How precisely and how quickly Selectio answers those questions weighs on every commercial opportunity as much as the product itself.

The problem

Selectio works with a range of more than 20,000 SKUs from multiple factories. Product information arrived in spreadsheets and images with no common standard, each supplier with its own columns, naming and language. Turning that material into something a buyer could browse was manual work repeated with every shipment of data: interpreting the files, matching each image to its product, translating and adapting descriptions, loading them into different tools, assembling the catalog and preparing quotes separately.

What clients received was the result of that effort, a set of hand-built spreadsheets. Every new product added hours of data entry, quotes took time, and the presentation fell short of the standard expected by the chains Selectio sells to. With a range of this size the model had a clear ceiling, because growing the catalog meant growing internal workload in the same proportion.

The solution

Together with Selectio, we designed a B2B platform built around the way it sells: an exclusive, login-only catalog for business clients, where each buyer signs in, searches thousands of products, builds an order and gets a quote.

The core of the solution sits at the point where data comes in. Instead of describing each product by hand, an AI cataloging tool works from the product image and generates the description, title, categories, attributes and metadata, along with translations adapted to each region. Spreadsheets and images from every factory become structured, consistent data under a single standard across the whole catalog, and adding a new product no longer depends on writing it up manually.

The buying experience rests on that organized data. An AI assistant built into the site and specialized in Selectio's catalog helps each buyer find products and suggests alternatives among thousands of options, while advanced search, brands, categories and themed collections structure navigation. Quoting covers what matters in an import operation: the volume in cubic meters (CBM) of each product and of the full order, current promotions and discounts, and quote export with saved changes, in the client's own view.

The platform also syncs products with Selectio's management system, so whatever is loaded or updated in the catalog reaches its back office without being entered twice. Since going live it has kept evolving, taking in new batches of products from the factories as they arrive and adding features as the commercial operation calls for them.

Results

• A digital catalog for a range of over 20,000 SKUs. What used to circulate as hand-built spreadsheets became a single, structured catalog with exclusive access for Selectio's clients.

• AI cataloging from an image. Description, title, categories, attributes, metadata and regional translations are generated automatically for every product.

• One data standard. Information from different factories is normalized under the same criteria, ready to search, quote and sync.

• Quotes built for importing. CBM volume calculation, promotions and export in the client's view are part of the same flow.

• A specialized buying assistant. Buyers explore the catalog with an AI assistant that knows Selectio's products.

• Catalog and management connected. Products sync with the company's management system, with no duplicate data entry.

Conclusion

A catalog of this size, fed by factories that share no standard at all, cannot be brought under control by adding hours of data entry. We solved it by placing artificial intelligence exactly where the operation was getting stuck, the description of each product, and building around it a platform designed for the way a large retail chain buys. On that foundation, Selectio can add factories, products and clients without its team having to grow at the same pace, and the platform keeps evolving through every stage of that growth.