Introduction
At the end of 2025, the Deutsche Bundesbank still had 16,799 domestic branches. That is six per cent fewer than in the previous year, following a decline of 8.4 per cent in 2024. The advisers who remain are expected to be available, conduct consultations, prepare appointments and fulfil extensive documentation requirements. As a result, the time available for advice is becoming scarcer and more valuable.
The first part of the series focused on accessibility for customers. This part shifts the perspective: AI voice agents also take the pressure off the advisers who make the biggest difference for banks. Three use cases illustrate where this has an impact in everyday practice.
Advisory time is becoming scarce: branch closures, skills shortages, documentation requirements
Face-to-face advice remains one of the most important success factors for banks, savings banks and building societies. Many institutions face the same challenges: fewer branches, a growing shortage of skilled staff, high regulatory requirements and customers who expect rapid response times.
For advisers, this creates a constant balancing act. They are expected to be available, provide advice, prepare for appointments, look after customers and, at the same time, keep records. More on the role of employees: Employee Experience.
AI Voice Agents in Banking: Answering calls whilst the adviser is in a meeting
The adviser is in a meeting with a customer. At that very moment, another customer tries to get hold of them. The phone rings, and the voicemail kicks in. The customer leaves a message and waits for a call-back, or tries again later. The result: missed contacts, repeated follow-ups and additional effort spent on call-backs.
The AI Voice Agent takes the call:
- It identifies the customer.
- It records the enquiry.
- It answers simple, standard questions straight away.
- It prioritises the urgency.
- It suggests a time for a callback.
Process: The agent greets the customer and clarifies what the matter is: a request for a callback, an enquiry about an appointment, a product-related issue or an urgent service request. It records who is calling, the reason for the call, by when the customer expects a response, and whether the enquiry can be resolved immediately or whether the adviser should call back. Instead of an anonymous voicemail message, the adviser receives a prepared call-back note detailing the enquiry, its urgency and the next step. Calendars or CRM systems contain all relevant information in a structured format.
A missed call is turned into a prepared customer contact.
Preparing for appointments: Clients arrive with all their documents
A consultation is due to take place. Often, documents are missing, appointments have to be rescheduled, or clients turn up unprepared. The adviser spends time organising rather than advising.
The AI Voice Agent proactively contacts the client:
- It arranges or reschedules appointments.
- It sends reminders about upcoming consultations.
- It requests any missing documents.
- It prepares the client for the consultation.
Process: A few days before the appointment, the agent sends a reminder to the client. It asks whether the appointment is still suitable, whether there are any outstanding questions and whether all documents are to hand. If any documents are missing, it specifies exactly what is still required and by when. If the appointment is no longer convenient, it offers alternatives or takes a request for a callback. Even before the meeting, the adviser can see whether the client is prepared, which documents are missing and whether the preliminary contact has revealed any issues.
The client arrives better prepared, and the adviser can focus on the content.
Documentation after the client meeting: Drafting rather than an evening spent at your desk
For many advisers, the administrative work begins after the meeting: documenting the results of the meeting, updating the CRM system, and meeting regulatory requirements. This often involves additional working hours, sometimes extending beyond normal working hours.
Immediately after the meeting, the adviser dictates the key findings into their phone. The AI Voice Agent:
- creates structured meeting notes,
- prepares CRM entries,
- extracts relevant information,
- and assists with the preparation of advisory and suitability documentation.
Process: The adviser describes what was discussed, the decision the client has made, which documents are missing and what the next steps are. The agent uses this to create a structured draft for the meeting notes, a to-do list and a CRM entry. The adviser reviews the draft, adds details and approves the documentation. Professional responsibility remains with the adviser, whilst the manual effort involved in taking notes, follow-up work and system maintenance is significantly reduced.
The obligation to document remains, but the manual work involved in documentation is considerably reduced.
Prerequisites: What is needed to reduce the workload on advisers using voice agents
Technology is rarely the bottleneck. In the CX-TOM study by the Lucerne University of Applied Sciences and Arts (2026, 43 banks), 74 per cent cited human resistance as the biggest hurdle with AI agents, followed by technical problems at 44 per cent. For Use Case 3, this means: the adviser checks and approves, and this ability to assess does not arise of its own accord.
Data integration is equally crucial. The banks in the study rated their real-time capability regarding customer context at an average of 2.44 out of 5. During the conversation, the customer must also be informed that they are speaking to an AI. Start with a use case and a clear success criterion, such as the proportion of missed calls.
Conclusion
AI voice agents do not replace advisory services; they create space for them. They answer customer calls immediately, take a proactive approach where required, coordinate callbacks and appointments, and draft call records. This frees up your advisers’ time for face-to-face conversations. Responsibility for advisory services and documentation remains with them: the agent drafts the records, whilst the adviser checks and approves them.
Experience the three use cases live.
The AI Voice Agent handles calls, coordinates callbacks, schedules appointments, and creates conversation summaries.
See msg.Hello in action: request a demo or join our free webinar on December 1.
