abstract banner


Helping a bank reduce cost-to-serve

We developed conversational agents to deal with requests, reducing time spent on credit card queries by 20%


A large international bank




Using chatbot and RPA to reduce cost-to-serve

  • Client challenge
  • Benefits to client
  • Approach
  • KPMG insights
Client challenge

Increasing your credit card limit may seem like a task that involves a great deal of analysis and consideration on the part of bank employees. You may imagine a bank employee reading through your statements, pondering whether to approve your request but in reality the process is more straightforward than you may think. There is a series of simple questions to be asked, limited answers to which determine the outcome, as well as some security measures to be fulfilled. There is little human intervention involved.


We decided to think big but start small, implementing the process step by step while getting buy-in from the corporate finance and IT teams in building our vision


Similarly when IT helpdesk employees make decisions on increasing the size of the mail box and personal disk space, their decision is based on clear, concise criteria with limited answers. Permission granting and right to install for applications can also be approved in this way.

Our client is a large international bank with a substantial number of staff. With a huge number of employees, the IT helpdesk can come under pressure. A high volume of requests is likely to slow down the level of productivity in the department, leading to operational inefficiencies. In this case, our client realized that these transactional helpdesk issues that were leading to such inefficiencies.

The processes were time consuming but with limited input and engagement from the agents. By automating these processes, employees could spend more time on more sophisticated tasks that add value to the company. Ultimately this could make the organization more efficient and improve the customer experience.

Benefits to client

We helped the client develop conversational agents to deal with credit card FAQs and credit card limit increases in addition to an internal chatbot to deal with the IT helpdesk queries. The benefits of automation soon became clear.

The automation of the credit card queries yielded a very positive result, reducing time spent on credit card queries by 20%.

By automating the process by which employees sought and received permissions associated with applications or increase their mail box size, employees who had generally carried out these tasks were free to focus on more strategic tasks that added value for the bank.

Between reducing the cost to serve and freeing up staff to work on tasks that add value, the automation of these activities yielded great results. .


This was one of the first projects undertaken by the bank in this area so it was very important that we get the basics right. Defining the bot and what we wanted the bot to do was a fundamental first step.

We used our experience in building our own internal chatbot to help guide the process. This learning process meant that we had a clear understanding of the goals and requirements that needed to be set at the start of the process in order to enable future success in the project.

This included identifying the key players in the client organization who would be involved in helping facilitating this automation journey both internally and externally. The considerations and requirements are also different when implementing a chat bot for internal vs external use.

The project also required assistance in choosing which IBM model/system to use. This involved analyzing a number of factors, including the short, medium and long term benefits of each approach.

The project had many layers – not only did it include technical implementation and an architecture component but it also required assistance in terms of ensuring compliance with local legal and regulatory constraints. This is an area in which our business acumen and knowledge of the local regulatory landscape was hugely helpful.

Although the challenge may seem purely technical, our business knowledge underpinned each stage of the process.

KPMG insights

A key aspect of this project was our ability to leverage our alliance with IBM. Having partnered with IBM on over 210 past engagements, we were confident that the combination of their leading edge technology and our deep regulatory, financial and operations knowledge would serve our client well and help facilitate their transition to conversional agents.

Another factor that contributed to the success of the project was the structured approach to implementing and executing the project. We decided to think big but start small, implementing the process step by step while getting buy-in from the corporate finance and IT teams in building our vision. The commitment and active participation of different teams at each stage meant that we could identify problems from each perspective early on, rather than finding ourselves trying to solve issues later on in the process because we moved too fast, without consulting those impacted by the project.

KPMG complies with the auditor independence rules of the AICPA, SEC, PCAOB and DOL. As a result, some services described herein may not be available to our audit clients. KPMG audit clients should check with their respective lead audit partner for more information.