Chat interaction on an iPhone between a customer and Agibank’s agent about the status of a loan request.
Agibank logo

Agibank answers 22,500 loan status inquiries per month with Agentforce

Agentic loan support delivers instant updates on WhatsApp, handling routine inquiries autonomously while routing higher-risk cases to service reps.

September 4, 2026

The results

75 %
deflection rate with Agentforce
22.5K
loan inquiries autonomously handled per month
245K
loan inquiries autonomously handled since launch

350,000 inquiries every month

Agibank is a Brazilian bank that has spent more than 25 years making banking more accessible for retirees, pensioners, and public sector employees. As the bank set out to grow their customer base by 40% without increasing operating costs, they needed a way to prepare for the additional service demand that expansion would bring.

Loan proposals were a natural place to start. Agibank's team of over 200 customer service reps already handled more than 350,000 WhatsApp inquiries every month, many of which involved customers asking routine questions about loan proposals.

Automating those high-volume requests could help Agibank absorb much of the expected growth while keeping service fast, personal, and focused on reinforcing customer trust.

24/7 loan status assistance

Agibank turned to Agentforce to meet that growing demand and automate one of their most common customer requests — while also keeping service fast and personal. Powered by customer data in Agentforce Service, their loan status agent gives instant updates on loan proposals, helping Agibank scale service without increasing fixed costs.

“We want to simplify the first contact experience and make service navigation more intuitive and conversational for customers,” said Akira Vargas Morishita, Service Agents Leader.

For customers, the experience is simple and begins the moment a customer reaches out on WhatsApp.

Agentforce greets them in Portuguese and asks how it can help. The customer can check the status of their loan proposal and get an immediate answer instead of waiting to speak with a service rep.

The agent also identifies opportunities to recommend additional services and will soon send outbound messages — like proposal approval notifications — instantly, making every interaction faster and smoother.

If the agent detects a compliance or fraud risk, it automatically escalates the interaction to a rep for human expertise and oversight.

Routed correctly, resolved quickly

When a customer starts a chat on WhatsApp, Agibank’s superagent greets them and asks for their Brazilian identification number (CPF) to verify their identity. The superagent triggers an action mapped to an omni flow, which checks the CPF against customer data in Agentforce Service.

Once verified, the superagent asks how it can help and interprets the customer’s natural-language reply. If the customer asks about a loan application, it uses an omni flow to bring the Loan Status agent into the conversation.

The agent then searches for open opportunities tied to the customer and turns information from the Order Object into a clear, conversational status update. It also logs a summary of the interaction in the customer’s Agentforce Service profile.

When a customer expresses interest in a new opportunity, Agentforce checks business hours to determine the next step: During business hours, it routes the conversation to the appropriate service queue. After hours, it explains when and how to connect with a sales rep.

Agibank uses Tableau Next to monitor the agent’s performance and track volume, customer intent, resolution outcomes, routing accuracy, and handoffs. These insights help teams measure performance and identify opportunities to improve the agent over time.

Proven value in three weeks

Within just three weeks of launch, Agibank's loan status agent was already delivering measurable value, handling 22,500 customer inquiries every month with a 75% deflection rate.

Building on that momentum, Agibank expanded the agent's role beyond loan status updates. It now automatically sends loan approval messages and presents relevant upsell opportunities during customer conversations, extending the value of every interaction.

“Our principal objective is to reduce effort for customers and provide fast, clear answers through automation, while preserving a welcoming, human tone and not being robotic,” said Morishita.

By removing friction from everyday interactions, Agibank is supporting growth without adding headcount — all while delivering the fast, personal service customers expect.

Why Agibank chose Salesforce 

Icon of a gear with a checkmark in the middle.

Salesforce Professional Services 

Starting with the loan status agent, Salesforce Professional Services helped Agibank design their first AI agents, establish governance, and put guardrails in place. The collaboration took place over six months and gave Agibank's teams the skills and confidence to maintain and improve agents on their own.

Icon representing low-code development tools and automated workflows.

Low-code tools

With Agentforce Builder's low-code tools, Agibank quickly launched the loan status agent. Starting with a high-volume, low-complexity challenge gave the bank a fast way to prove the value of agentic AI.

Icon representing a "trusted view of every customer," featuring a person silhouette combined with security, visibility, and growth symbols.

One trusted view of every customer 

Agibank's hybrid model depends on consistent customer information across digital and in-person channels. Agentforce Service gives service teams a single view of each customer while meeting the security and compliance requirements of the financial industry.

Agibank logo

About the company

Agibank is a Brazilian bank combining digital banking with in-person service, offering loans, credit cards, insurance, and investments.

Get started