Tecban took a structured, company-wide approach to introducing AI, rather than running isolated pilots.

They focused on two main fronts:

• Business productivity with Microsoft 365 Copilot and Copilot Chat, embedding AI directly into tools employees already used every day (like Office 365).
• Software development productivity with GitHub Copilot, supporting development and testing teams.

This approach helped Tecban:

• Deliver 98 AI use cases, with 35 already completed.
• Onboard more than 3,000 users overall and reach over 1,800 Copilot Chat users.
• Capture around 12,600 hours of assistance, estimated at roughly R$ 946,000 in value.
• Report approximately 5,808 hours per year in productivity gains in 11 Microsoft 365 Copilot cases, estimated at about R$ 800,000.

On the development side, Tecban estimated:

• About 25–30% productivity gains for code creation with GitHub Copilot.
• Approximately 30% productivity gains in development overall and 14% in testing.

In day-to-day business work, tasks that used to take hours now take minutes or seconds. For example:

• An analysis that took 4 hours and 47 minutes is now completed in 2 minutes and 35 seconds.
• Another process dropped from 1 hour and 35 minutes to just 52 seconds.

By integrating Copilot into existing workflows and measuring outcomes, Tecban was able to increase productivity by up to 30–35% in key areas while keeping governance and control in place.

Tecban operates Banco24Horas, a critical financial infrastructure with about 24,000 self-service banking devices across Brazil. In this context, availability, traceability, and control are non‑negotiable.

To scale AI safely, Tecban moved from ad hoc experiments to a formal corporate AI program with clear governance. Key elements included:

• A structured intake and evaluation flow for every AI use case: demand registration, risk assessment, data sensitivity analysis, technical validation, and only then production rollout.
• Early involvement of HR, risk, process, security, and business teams to build digital literacy and define responsible-use policies.
• A focus on avoiding shadow IT and fragmented tools by standardizing on the Microsoft ecosystem already in use.

By choosing Microsoft 365 Copilot, Copilot Chat, Copilot Studio, and GitHub Copilot, Tecban could:

• Keep identity, data, security, and auditing within a consolidated platform.
• Embed AI directly into existing tools (like Office 365) instead of introducing disconnected external solutions.
• Maintain centralized control over how AI is accessed and used across the organization.

This model allowed Tecban to rethink how AI is deployed: not as one-off experiments, but as a managed capability that balances speed, productivity, and regulatory compliance in a critical financial environment.

Tecban focused on making AI tangible in everyday work by building purpose-built agents and use cases rather than abstract demos.

Some of the key practical applications include:

• Norbot: An AI agent for the commercial and product teams that automates much of the internal pre‑sales flow, helping teams prepare faster and more consistently.
• Norminha: An agent focused on compliance standards, supporting teams that need to follow strict regulatory rules.
• Clara: An assistant for the auditing team, helping with document review and analysis.
• Conecta RH: An HR-focused agent that supports recruitment, selection, and internal services, improving response times for employees.
• Field technical assistant: An AI helper for operators who service Banco24Horas devices across Brazil, providing guidance and information in the field.

These agents were built using Copilot Studio and integrated into Tecban’s existing Microsoft environment, so employees could access them within familiar tools.

The impact was clear in daily operations:

• Repetitive tasks were offloaded to AI, freeing teams to focus on higher‑value analysis and decision-making.
• Processes that once took hours were reduced to minutes or seconds, with measured time savings and cost benefits.
• AI shifted from being a “future concept” to a working partner embedded in each team’s routine.

Looking ahead, Tecban plans to extend this model to more complex, business‑critical processes, evolving from simple assistants to more continuous, specialized AI agents that act as true collaborators in core operations.