Argentine bank Banco Hipotecario has deployed two AI agents built with Microsoft Copilot Studio following a two-year migration to cloud-based Dynamics 365, leading to a 114 percent increase in Net Promoter Score. The bank, which specializes in mortgage finance, personal loans, savings, and investment products, moved from an on-premises deployment of Dynamics 365 to Dynamics 365 Online between 2023 and 2025.
The cloud migration gave the bank access to Dynamics 365 Sales, Dynamics 365 Customer Service, and a modern analytics stack that includes Databricks, Azure Data Lake, and Power BI. The bank replaced older SQL console reporting with these tools, enabling branch-level reporting for the first time. Azure Application Insights monitors systems in real time, and Azure Key Vault handles secure credential management. The development team uses Azure DevOps and GitHub Copilot under bank-provided licenses.
"Moving to Dynamics 365 Online gave us the foundation to modernise how we serve customers and support employees," said Rodrigo Espinosa, head of customer relationship management at Banco Hipotecario. "With Copilot Studio, we are turning that foundation into faster answers, better experiences, and a more agile bank."
MAIA and DynaBot: Two different AI agents
The bank launched MAIA in 2025 to give employees instant access to operational, IT, and product information through a single conversational interface. Before MAIA, workers had to submit helpdesk tickets and wait days for answers found across PDFs and the bank's website. "All the bank's employees need better tools to do their work every day, and we used Copilot Studio to create MAIA, an agent that automates information and tasks and brings everything together into one channel," said Pablo Ruggieri, technical lead at Banco Hipotecario.
The second agent, DynaBot, is still in pre-launch testing. It captures customer information and automatically creates incidents in Dataverse. Internal tests show DynaBot can reduce average case creation time by 80 percent compared to manual processes. The agent handles debit-card discrepancies, complaints, product requests, and account issues. Customers will reach DynaBot through WhatsApp and Microsoft Teams, while customer-service agents manage incidents through a shared workflow. "Dyna will make the entire support process much faster for our customers, and enable us to give them more attention," said Ruggieri. "That will help us improve our customer satisfaction scores."
For anyone working in AI for Customer Support, this case shows real, measurable improvements from agent deployment options. The bank recorded a 182 percent increase in WhatsApp operator support, a 150 percent increase in phone support, and a 9 percent increase in branch support between June and September 2025. The overall Net Promoter Score jumped 114 percent during that period.
"With the bots we built using Copilot Studio, we have seen improvements for our customers using WhatsApp, and as a result, the human attention we give them is better than before," said Ruggieri.
Why this matters for Customer Support professionals
Banco Hipotecario's deployment shows that AI agents can deliver both internal efficiency gains and external customer satisfaction improvements. The 80 percent reduction in case creation time for DynaBot, coupled with the multichannel support improvements, translates directly to faster responses for customers, and agents' ability to focus on complex issues rather than data entry.
For teams exploring similar tools, the Microsoft AI courses offered by Complete AITraining explain the platform behind these agents, giving support professionals a practical path to build and deploy similar systems. Customer support teams looking to automate routine interactions while preserving quality human attention - exactly what Banco Hipotecario accomplished - should study this model before launching their own agent rollouts.
Your membership also unlocks:
