Background and Motivation
Grupo Reyes, a family-owned conglomerate behind the Reyma brand, has been adapting to Mexico’s market cycles for 55 years, transitioning from manufacturing, construction, hospitality, and now, a decisive shift towards artificial intelligence (AI). Abraham Sevillano, Director of iNN Innovation and Businesses at Grupo Reyes, explains that the need for AI implementation arose from exponential growth and diversification over the past decade, generating a vast amount of structured and unstructured data.
Cultural Shift and Agent Conversational Implementation
The first step was a cultural shift, moving away from complex metric navigation for executives. Instead of forcing them to delve into intricate data, Grupo Reyes’ team opted for a conversational agent inspired by ChatGPT, breaking the barrier of understanding data in natural language.
“We created our own agent, and you can talk to it in natural language. Behind that simple response is a full analysis,” said Sevillano.
The goal was to transition from traditional dashboards that show what’s happening to AI models predicting what could happen.
AI Implementation with Dell and Nvidia
Dell and Nvidia collaborated with Grupo Reyes to design a viable, contained, and formative AI implementation. The focus was on developing internal talent rather than purchasing an out-of-the-box solution.
“The first step was establishing a trust-based communication. Grupo Reyes wanted to develop internal talent. They didn’t want to buy a ready-made solution,” said Kurt Yáñez, Dell Technologies México’s AI Business Development Leader.
This approach allowed Grupo Reyes to start with an experimental budget, accompany the development team, and convert proof-of-concept trials into solutions with tangible returns.
Simultaneously, Nvidia observed something uncommon in Mexico: genuine work with generative AI (IAG).
“Few companies are truly developing IAG concepts; they’re using our programs like Nimo, Nimo Cloud to build their own language models,” said Marcio Aguiar, Nvidia’s Enterprise Director for Latin America.
What Grupo Reyes Did with AI
- Centralized data: Grupo Reyes built a data lake to unify structured and unstructured data sources, feeding models until the optimal fit was found.
- Selected mundane but critical use cases: Instead of chasing demo shows, they tackled operational friction that could be measured. For example, analyzing thousands of support tickets to identify trends by business sector (plastics, construction, hospitality) and prioritize attention based on urgency.
- Ensured adoption: They created an internal marketing team to change the perception of the area and communicate clearly with executives and users, breaking down the technological complexity barrier.
Infrastructure: Start Small, Scale Sensibly
The technical design of this AI implementation followed the “quickly prove value and grow on demand” approach. Grupo Reyes began with a Dell PowerEdge R760X server with two GPUs, sufficient for initial use cases based on unified information and ticket analysis.
“If I launch the agent to 4,000 users (with administrative access), my computing power will be minimized,” said Sevillano.
AI Factory
Dell and Nvidia’s support structured this approach within the AI Factory framework, which involves identifying business cases, defining models and data (cleaning, governance, access), and only then landing the technological architecture.
“We launched it at Dell Technologies World, a framework that helps implement use cases in shorter timeframes,” recalled Juan Francisco Aguilar, Dell’s General Manager in Mexico.
Marcio Aguiar added that Nvidia’s software suite aims to accelerate time-to-market, not just sell hardware, and allows building the “AI factory” by scaling with demand, without overinvesting upfront.
ROI of AI?
Unlike those promising instant returns, Grupo Reyes separated the AI implementation process evaluation into two phases: a qualitative phase with user and executive experience surveys, followed by a financial phase once models are deployed directly in production and procurement.
“We’ve done well in the qualitative phase… the next phase is financial, as we deploy models directly into production and procurement,” said Sevillano.
The time to achieve initial results was reasonable: six months for technical aspects and around a year for the user experience to be felt within the company.
People-Centric Approach
Sevillano emphasizes that AI is not a pretext for labor reduction but a talent accelerator.
“AI won’t replace any collaborator; it will supercharge their decision-making and empower them,” he said.
The strategy combines intergenerational mentoring, transmitting corporate culture to younger teams, with aspirational projects, aligned benefits, and an AI career narrative within the group.
“We’re not thinking about cutting jobs; we aim to improve productivity… that’s our virtuous cycle: better quality of life, better decisions, higher profitability,” he said.
This vision aligns with Nvidia’s diagnosis: the challenge isn’t replacing people but learning to work with AI.
“The computer will work side-by-side, bringing highly relevant information to your business,” said Aguiar.
Scope
Grupo Reyes’ case doesn’t emerge in a vacuum. The holding has a national presence with plants in Mexico City, Sahagún, León (corporate), Monterrey, Merida, Nogales, and a distribution center in Phoenix. They also have 118 business points across other units.
Reyma, their most recognized company, dominates the single-use disposable market. Recently, they expanded into construction and hospitality sectors (salons, restaurants, and hotels).
The generational shift also plays a role. “The owner’s children are now entering and have no access to any system; only to our agent and dashboards,” said Sevillano.
This practice serves as an indicator of AI adoption: new leadership no longer wants menus and manuals; they want to ask questions and make decisions.
What’s Next: Production, Procurement, and Data Security
After a year of learning, Grupo Reyes plans to take models to hard processes (production lines, supply chain) and open the agent to thousands of users with business-specific roles and permissions.
This requires more graphics processing units (GPUs) and a finer data security and governance architecture. The trend is clear: many organizations are requesting “mass production,” repatriating cloud loads to on-premises due to cost efficiency and risk control in AI usage.
This appetite isn’t exclusive to Grupo Reyes. Dell and Nvidia see manufacturing, finance, and transportation in the Bajío as AI technology adoption leaders.
Grupo Reyes’ case doesn’t showcase a futuristic AI laboratory but strategic discipline applied to a Mexican conglomerate operating across various verticals.
