Cemex creates a copilot to cement its foundation with customers

When multinational building materials company Cemex recognized the need for an AI tool to enhance the agility of their sales force and enable them to tailor solutions to their customers’ needs in a matter of minutes, they also understood that efficient human collaboration within the company was crucial to the success of their plan.

The result of this combination was a sales copilot called Technical Xpert, an AI-assistant running on Microsoft Azure OpenAI Service with GPT 4.0 and found as a chat window in Teams.

Spurred by a need in Mexico operations and supported by Central technology area to create an AI-powered tool that would increase efficiencies in the sales of its products, and even change the way the company offers holistic solutions to its customers,  Noel Martínez Pérez and Alan Cortez joined forces to build a solution that would help the sales team quickly navigate Cemex’s vast array of products, from traditional concretes and cement to its new suite of sustainable construction materials.

“I think we are in a time when we need to integrate all these types of tools into our daily processes, because although it is something that has been working well for a long time, I think that with this we can take it to the next level,” says Cortez, head of commercial information at Cemex Mexico. “And that’s exactly what we’re looking for, to think out of the box and to be able to evolve our processes.”

 “Many times, hard data provides us with information that, as data scientists, we believe is insightful,” says Martínez, a data scientist for Cemex. “But those hard data, those hard numbers, don’t always reflect what the customer needs.”  Which is why the synergies between both departments were crucial to come up with the right copilot.

Over the development period of the AI tool, Martínez and Cortez went back and forth, testing different elements and accentuating several features to provide the best possible product.

Two men sit talking next to one another each at their own computer desk
Alan Cortez and Noel Martinez Perez used their skill sets in sales and data science to collaborate on the Technical Xpert tool for Cemex. Image by Octavio Hoyos.

“The kinds of details that the bot should look at, I think, come out of that synergy between Alan and our team, where they tell us, which functionalities of the bot would be useful, and based on those user recommendations, we programmed and deployed the tool to do just that” Martínez says. “We polished the tool until it was ready.”

The outcome of these efforts, alongside their partnership with Microsoft, is a solution that aligns with Cemex’s digital transformation efforts. Technical Xpert is a copilot embedded that allows Cemex’s sales agents to use AI to pull up all pertinent product information – almost like an instantaneous online brochure – and provide the best possible materials for customers in a fraction of the time.

The tool recently launched in Mexico, and sales agents are beginning to incorporate it into their engagement with customers, with game-changing results.

“This new tool is sending us information much faster, so we can have the conversations with customers and at the same time asking, internally, technical questions: What can I offer to this customer based on what he’s requiring from me?” says Cristhian Alday, a Cemex commercial advisor located in La Paz, Baja California. “For me, reaching this type of information in the past would take me to knock at least on four people’s doors and the response time would have been maybe, I don’t know, in an hour. And here it is in nine seconds.”

Cemex relies on a network of sales agents throughout Mexico to provide potential customers with the solutions that work best for a particular project. But with so many different choices now available, even the most knowledgeable agents could get stuck in their attempts to recommend roller-compacted concrete or masonry mortar cement.

A man stands with arms folded in an office
Fausto Sosa, vice president of information technology at Cemex, oversaw the development of Technical Xpert. Image by Cemex.

“Sometimes it is very hard for the sales force to know all the different ranges of products and the correct technical characteristics for them,” says Fausto Sosa, vice president of information technology at Cemex. “We are constantly releasing new products that are more sustainable and with different characteristics. So, it becomes a little complex to understand all the details.”

That led to Cemex Mexico president suggestion of building a tool to help the sales force become more agile and effective in its recommendations.

“He wanted to have a sales force with relevant information because they are not in offices, they are in the field,” says José Luis de Apodaca, global data science director for Cemex. “It’s a role that usually doesn’t sit at a desk and go over dashboards. So, we needed a very efficient and available tool.”

As part of that change, Cemex looked to Microsoft and the power of AI for help. How could it create a nimble, knowledgeable tool that empowered its sales agents to make the most valuable and efficient product recommendations for its customers? With each large construction project, sales agents needed to put together a full set of materials required, also providing options to account for environmental factors, ground composition and cost.

For the Microsoft team involved in helping create the copilot, the challenge was ensuring that Cemex had an architecture that both met their needs and was highly secure. The back and forth ensured that the design accounted for all variables and could be flexible as new products entered Cemex’s system.

This customized copilot had to be built and trained with the correct information over four months. That required giving the tool hundreds of pages of documents and manuals with information on the myriad products Cemex offers to customers. It also had to be easy to update, since new developments in materials would allow sales agents to have the latest products available to suggest to customers.

Latest articles


Related articles

Leave a reply

Please enter your comment!
Please enter your name here