Tecban increases up to 35% productivity with Microsoft 365 Copilot + GitHub Copilot

Critical infrastructure, national scale, and the need to make AI work with control.

Tecban operates Banco24Horas, the largest independent self-service banking network in the world. With approximately 24,000 devices deployed across Brazil, the operation connects banks, retailers, and millions of Brazilians. At this kind of scale and reach, availability, traceability, and control aren't competitive differentiators, they are baseline requirements.

That's the context in which artificial intelligence entered Tecban's agenda. AI adoption became part of a broader organizational transformation at a company that already places operational efficiency at the center of its strategy. But in a highly regulated financial environment, scaling AI wasn't just about rolling out new tools, it meant turning productivity into business results without compromising governance, security, or compliance, while avoiding technological fragmentation and the risk of ungoverned AI adoption outside the corporate environment.

For Tecban, the starting point wasn't experimentation for its own sake — it was operational. The goal was clear: gain productivity without compromising governance, security, or compliance in an operation that's critical to Brazil's financial system. That balance between speed and control has guided the entire journey.

"To keep moving forward, AI can't remain a one-off effort, it has to become part of the day-to-day business and processes of the organization," says Vanessa Oliveira Ferreira, IT Superintendent at Tecban.

From one-off initiatives to a corporate path for AI

The first uses of AI emerged within technical teams, particularly around software development. But Tecban quickly recognized that the topic would need to evolve from isolated experiments into a structured program with clear criteria to scale safely. In a regulated environment like financial services, letting adoption happen organically wasn't enough, the company needed a fast, secure corporate path to avoid fragmentation, parallel tools, and the risk of shadow IT.

Before scaling individual cases, Tecban needed to build clarity around AI's potential in day-to-day work, structure how new initiatives were brought in, and establish measurement criteria so projects could move forward with defined value and clear guardrails. More than simply expanding the use of technology, the company set out to establish a consistent methodology for AI deployment, ensuring that every new case followed defined stages of evaluation, governance, and production rollout. From that point on, cases moved through a structured flow: demand registration, risk assessment, data sensitivity analysis, and technical validation before going live.

The strategy was built on two simultaneous fronts: GitHub Copilot for development teams, and Microsoft 365 Copilot and Copilot Chat to bring AI into the daily routines of business areas. "We started on two fronts: GitHub Copilot to boost productivity in development, and Microsoft 365 Copilot and Copilot Chat to support the business areas," explains Vanessa Oliveira Ferreira, IT Superintendent at Tecban. This combination allowed Tecban to start with cases that delivered immediate practical value, while laying the foundation for broader AI expansion across the organization.

At the same time, adoption was not treated only as a technology rollout. The company involved HR, risks, processes, security and the business areas from the beginning, both to build literacy and digital culture and to sustain a clear policy of responsible use. This collective work helped form the structured basis that allowed us to scale the journey more consistently.

The value of scaling within the Microsoft ecosystem

In structuring this journey, Tecban chose to expand AI within the Microsoft ecosystem that already powered its digital operation. The goal wasn't just to accelerate adoption, it was to reduce technology fragmentation, avoid new data silos, and lower the risk of shadow IT by keeping identity, data, security, and auditing within a foundation that was already consolidated.

With Microsoft 365 Copilot, Copilot Chat, Copilot Studio, and GitHub Copilot, Tecban began embedding AI directly into the tools teams were already using, with technological consistency and centralized control.

"The main point for us was integration with the ecosystem we were already using. With Copilot, that experience happens right inside Office 365," notes Luiz Fernando Ribeiro Lopes, Executive Manager of Digital Platforms and Integrations at Tecban.

This approach allowed the company to move forward with stronger governance, without turning the pursuit of productivity into a proliferation of disconnected external tools sitting outside the corporate environment.

From assistants to agents: making AI tangible in everyday work

With the governance foundation and adoption base in place, Tecban began turning AI into concrete applications across different areas. One of the clearest examples is Norbot, an agent built for the commercial and product team that automated much of the pre-sales flow inside the organization. It helped make the strategy tangible for teams and showed that AI could move beyond conversation into practical, hands-on roles in everyday work.

From there, the company expanded the model with agents purpose-built by function and area. Norminha was created for compliance standards, Clara supports the auditing team, and Conecta RH assists with recruitment, selection, and internal services. A field technical assistant was developed to support operators servicing Banco24Horas devices deployed across the country. Rather than using Copilot Studio in the abstract, Tecban set out to build agents with a clear purpose and real-world application. As Vanessa Oliveira Ferreira, IT Superintendent at Tecban, puts it: "When teams see the time they're no longer spending on repetitive tasks, AI stops being a promise and becomes real value, day in and day out."

Measurable impact on adoption, productivity, and value

As the journey matured, Tecban turned isolated initiatives into a structured corporate program. The company has logged 98 use cases, with 35 already completed, and has reached more than 1,800 Copilot Chat users across the organization.

Over the reported period, 12,600 hours of assistance were captured, equivalent to roughly R$946,000. In 11 of the 35 completed Microsoft 365 Copilot cases, business areas reported gains of approximately 5,808 hours per year, estimated at around R$800,000.

On the development side, the results were just as concrete. Tecban estimated a 30% productivity gain in development and 14% in testing, reinforcing GitHub Copilot's role as a technical accelerator within the broader strategy. The same momentum showed up in day-to-day activities: "An analysis that used to take 4 hours and 47 minutes is now done in 2 minutes and 35 seconds. In another case, it dropped from 1 hour and 35 minutes to 52 seconds," notes Fabio Viel dos Santos, Process Manager at Tecban.

In some flows, the reduction was even more dramatic. "For code creation, we estimate at least a 25% to 30% gain with GitHub Copilot," adds Bruno Bogaz Rodrigues, AI and Hyperautomation Manager at Tecban. More than just speeding things up, results like these showed teams that AI could absorb repetitive work and free up time for higher-value analysis.

The next step: AI that's more agentic, more scalable, and closer to critical processes

With the foundation in place, Tecban's ambition is now to expand the use of AI to more complex, more critical processes that are more connected to decision making. The logic is no longer just that of punctual assistants and moves towards agents with a more continuous, more specialized role and increasingly closer to the core of the business.

“Our ambition is to have agents acting as true working partners within each Tecban team,” concludes Vanessa Oliveira Ferreira. This expansion tends to happen supported by the same model that structured the initial phase: sponsorship of the business areas, participation of HR, processes and security, and close partnership with Microsoft to transform literacy into real application. For Tecban, this was the way to move from curiosity about AI to a scalable organizational capability.

Discover more about Tecban on Facebook, Instagram, LinkedIn, and YouTube.