Company file

Company / Enterprise technology

The Mexican Software Company That Made Nearshore a Strategy

Long before nearshoring became a boardroom buzzword, Softtek built an international IT business around proximity. Now the Mexican company is testing whether that human advantage can survive an agentic turn.

The cleverest thing Softtek sells may be the hours when everybody is awake. In 1997, as American corporations became comfortable sending software work halfway around the planet, the Monterrey company offered a different map. Put engineers in Mexico and other nearby countries, it argued, and a client in the United States could ask a question in the morning, discuss the answer before lunch and fix the problem while the context was still warm. Softtek gave that arrangement a name: nearshore.

Today the word has escaped the IT department. Executives use it for factories, supply chains and geopolitics. Softtek's version remains more specific. It is an operating model for building and running enterprise technology with teams close enough in time zone, geography and working culture to behave like colleagues rather than a distant ticket queue. That idea helped turn a small business founded in Monterrey in 1982 into a privately held software engineering group with more than 15,000 technologists and operations across the Americas, Europe and Asia.

The company does not arrive with a single must-have app. It arrives when a retailer's order system cannot keep up, when a bank needs to modernize old code without interrupting payments, when a manufacturer wants SAP to bend around a real supply chain, or when a sports club needs millions of fans to feel that its app, website and stadium belong to the same place. Softtek sells the design, build, integration, testing, security and long-term operation required to make those systems work.

1982Founded in Monterrey, Mexico
15K+Technology professionals
40+Years inside enterprise systems

The product is proximity

Outsourcing has always promised a triangle of benefits: more talent, lower expense and faster delivery. The triangle gets wobbly when coordination consumes the savings. Twelve-hour gaps stretch a small clarification into a two-day exchange. Requirements pass through account managers and come back altered. A late discovery becomes a late release. Softtek's distinction is to treat those frictions as engineering variables, not unavoidable facts.

Nearshore delivery does not guarantee good software. It does change the conditions under which good software can be made. Teams can join the same planning sessions, pair on difficult work and involve a client's security or product leaders while a decision is still reversible. Travel is shorter when a screen cannot replace a room. For North American buyers, Latin American engineers also provide a way to diversify beyond the huge India-centered suppliers without trying to recruit every skill in-house.

That places Softtek in an unusual middle. It is broader than a development boutique and can take responsibility for a full technology estate. Yet it is smaller and more regionally rooted than Accenture, Tata Consultancy Services or Cognizant. Globant, CI&T, Encora and other Latin American engineering firms contest the same proximity advantage, while clients can build their own captive centers. Softtek's answer is institutional memory: decades of delivery methods, platform specialists and reusable tools that a newly assembled team does not have on day one.

“It was a way to position Latin America not only on low cost, but on talent.”Blanca Treviño, co-founder and longtime chief executive

From billable hours to reusable intelligence

The awkward question for every technology-services company is what AI does to the timesheet. Clients will not pay the same price for work a machine makes dramatically faster. But they will pay for a provider that can make automation safe inside a messy, regulated organization. Softtek has spent more than a decade developing FRIDA, short for Framework for Intelligent Digital Automation, to occupy that space.

FRIDA is not presented as another chatbot or a boxed software suite. Softtek describes it as a delivery system: reusable agents, accelerators, client context, testing intelligence, controls and accumulated knowledge that sit across the lifecycle of software. One branch helps turn natural-language requirements into functional tests. Another helps generate documentation, scaffold code or convert legacy languages. Operations agents can assist with diagnosis and recommendations while people retain approval and accountability.

How the delivery machine compounds

01SpecifySenior teams define the work and the rules.
02ExecuteEngineers and agents build, test and operate.
03ValidatePeople review outputs, risk and business fit.
04RememberUseful context returns to the delivery system.

This changes the pitch from selling more effort to scaling outcomes. It also exposes the hard part. Enterprise AI must inherit permissions, audit trails, architecture constraints and business language. A generated test is useful only if it covers the requirement that matters. A modernization agent is useful only if the new code behaves like the old system where it should. Softtek's ISO/IEC 42001 certification, announced in November 2025, matters here as an operating signal: the company has put a formal management system around AI risk, monitoring and accountability.

Abstract Swiss-style composition of connected nodes, modular blocks and cloud forms
Everybody awake at once. The arcs are distance; the bright nodes are where a decision no longer has to wait overnight.

What clients actually buy

The catalog is wide because large systems rarely fail in neat categories. Softtek develops new digital products, integrates APIs and modernizes old applications. It migrates and operates cloud environments across AWS, Microsoft Azure and Google Cloud. It implements and manages enterprise platforms including SAP, Salesforce and ServiceNow. Its security practice covers identity, application and cloud controls, threat monitoring and incident response. Data teams build governed foundations for analytics and AI. Managed-services teams stay after launch to handle performance, incidents and change.

Customers span financial services, insurance, consumer goods, retail, automotive, healthcare, energy, telecom, technology, government and transportation. Many case studies keep the client unnamed, common in enterprise contracting, but the problems are recognizable. One Latin American department-store chain used a Softtek-built logistics platform to raise customer-service capacity by as much as 70 percent, double direct home deliveries and reduce first-year last-mile costs by 30 percent. A government revenue agency used incremental platform development to handle intense filing peaks and improve self-service. A consumer-goods engagement launched seven direct-to-consumer brands and reported a 44 percent revenue increase tied to faster e-commerce launches.

Where it can help: when an enterprise needs a durable product team, a legacy estate moved without drama, around-the-clock platform operations, automated quality engineering, or a nearby delivery center that can expand faster than internal hiring.

The most visible proof is sports. Softtek designed and developed Real Madrid's relaunched fan app and has worked on the club's main website and a dedicated Bernabéu portal. In March 2026, the company said active app users had grown more than 150 percent and the app held a 4.5-star store rating. The work connects live match coverage, video, player information, merchandise and tickets - the ordinary pieces of fandom made difficult by enormous traffic, languages and commercial stakes. Softtek also became the Dallas Mavericks' official technology sponsor in 2024, linking its engineering brand to a second large fan ecosystem.

“Softtek has been a true strategic partner, involved at every stage of the process.”José Manuel Peña Arroyo, CTO of Real Madrid C.F.

A services business under new pressure

Softtek makes money through consulting and engineering projects, dedicated teams, platform programs and recurring contracts to manage applications and infrastructure. The company is privately held, so it does not publish the quarterly detail that would show exactly how quickly that mix is changing. The strategic direction is visible anyway. It wants FRIDA to improve the economics of delivery while its people remain responsible for the result.

That is both sensible and difficult. Rivals have their own copilots, automation platforms and delivery centers. Cloud providers increasingly offer professional services around their ecosystems. Clients are building internal AI teams and asking every vendor to prove that productivity claims become lower cost, faster release cycles or fewer incidents. “AI-enabled” is no longer a differentiator by itself. Softtek has to show that its agents know the client's environment, that governance travels with the work and that nearshore teams resolve exceptions better than a remote factory.

Its latest alliances reveal the intended route. AWS recognized Softtek with a DevOps Competency in February 2026, after the airline edition of FRIDA ITA passed an AWS technical review. A July alliance with Enactor targets unified commerce for retailers in Mexico and Latin America. Real Madrid renewed its digital relationship in 2025. These partnerships put Softtek beside platforms and brands that already have distribution, while the company contributes integration, localization and long-run engineering.

The succession test

There is also a human transition. Blanca Treviño, a co-founder who served as chief executive for 26 years, is moving to executive president while remaining board chair. David Jiménez is scheduled to become global CEO on October 1, 2026. For a company whose identity is tied closely to Treviño and to the category she helped popularize, the split is significant. Jiménez inherits a global delivery organization just as AI is changing how its units are priced, staffed and measured.

The durable lesson in Softtek's story is not that close is always better than far. It is that the hidden costs of distance can become a product opportunity. In the 1990s, the hidden cost was a workday lost between question and answer. Softtek packaged overlapping hours, travel and regional talent into nearshore. Today the hidden cost is context lost between human teams and AI tools. FRIDA is an attempt to package that context into a governed system that remembers.

If it works, Softtek will not look like a company replacing engineers with software. It will look like a company making small, senior groups capable of carrying more responsibility. That would preserve the original nearshore promise: not merely cheaper labor nearby, but better conversations around consequential technology. The geography still matters. The quality of the handoff - between client and team, agent and engineer, old system and new one - matters more.

Keep exploring