Using gen AI to redefine the commercial banking front office

Not all use cases in a commercial bank are meaningful candidates for a gen AI uplift. Generally, use cases fall into three broad categories: front office, headquarters and back office.

Use case framework: Evaluating the client risk of various Gen AI use cases

For commercial banks, it’s important to keep in mind that the first use case will help stakeholders understand the benefits, risks and potential opportunities that are possible in their ecosystem.

When evaluating where to begin, keep in mind that pilot projects should have limited exposure and risk. Two dynamics that help prioritize use cases are:

• Level of direct client interaction – uses cases with lower direct client interaction allow the employee to be augmented by the technology, reducing the risk of incorrect information being shared with the client.
• Level of risk of negative client impacts – uses cases with lower risk of negative client impacts allow for early experimentation with AI.

Example gen AI use cases for commercial banking activities

The use cases in the bottom left quadrant accomplish two things: they reduce risk to the business and they allow RMs and front-office staff to experiment with tools.

Examples such as synthesizing customer information or creating key insights are common starting points to integrate a technology and productivity uplift. An interesting use case to explore from this quadrant is a front-office gen AI tool, the Augmented RM Dashboard.

The RM plays a crucial role in building a strong level of trust and partnership between the bank and its clients. RMs, however, often stray from focusing on the most important aspects of client relationship-building and more on the smaller, more time-consuming parts of the job. These challenges include staying updated on a vast array of products, negotiating loan terms using manual pricing models, conducting extensive market research and competitor analysis, and managing ad hoc responsibilities such as meeting notes and synthesizing key findings.

These challenges ultimately hinder an RM’s ability to remain client-centric and able to offer the vast array of assets and services their institution provides. Sacrificing either of these  damages the value proposition of a financial institution serving a commercial business.

Banks that get it right can better retain their clients.

How can we use gen AI to improve the RM experience?

The key differentiator between gen AI and automation tools is that gen AI is forward thinking and can be used for more than just reducing simple tasks. In fact, 66% of banks believe gen AI will enable greater productivity by automating targeted sales prospects and outreach for RMs.¹

When considering the dynamic between the RM and the customer, key differentiators include providing the right information at the right time in a way that is meaningful to the customer.

A tool that helps check off many boxes for the RM is the Augmented Relationship Manager Dashboard. This tool helps blend the virtual environment with real-time solutions, allowing RMs to provide a customer experience that is direct, tailored and differentiated.

Using voice recognition, the dashboard works as a “personal assistant” in the background of a virtual meeting, providing recommendations and next steps to the RM based on the conversations with the customer.

With a tool that can serve as a personal assistant to an RM and help provide real-time summaries, analysis and recommendations, an augmented RM can help most front-line activities.

The below table indicates tasks that an RM is responsible for on a day-to-day basis. Using an augmented RM dashboard, many of them can be addressed and achieved.

Augmented RM dashboard illustrative example: A commercial business is looking to grow their operations and wants to expand to an additional location in British Columbia. The Director of Finance has approached their RM to discuss the feasibility of a $2m loan. Below is their interaction and how an AI tool can support the RM.

When considering the implementation of a tool such as the augmented RM, it’s crucial to recognize that front-line staff often view these changes with skepticism, as they can disrupt their daily workflows. Consequently, tools advocated by head office and technology teams, such as Salesforce and other automation solutions, are frequently underutilized.

Banks should view gen AI as a strategic and enterprise-wide asset, so most decision-making regarding its purchasing and deployment should be a collaborative effort, not just the technology and IT functions. This shift in mindset can provide further direction to IT teams on how and where to focus development efforts.

Building key capabilities as a commercial bank

While investments and strategic buy-in play a key role, commercial banks must also consider maturing a set of key enabling enterprise capabilities to support their gen AI initiatives:

Our framework: Key enabling enterprise capabilities

BUSINESS ENABLEMENT & STRATEGY

• Include gen AI strategy as in the organization's broader IT strategy.
• Define industry use cases; define KPIs for measuring business impact, ROI and risk to support strategic action.
• Develop an intake model to manage use case proposals from ideation, evaluation and deployment.
• Help establish or enhance centres of excellence (COEs) to enable experimentation, innovation and adoption of gen AI in both the tech/IT and product teams.

DATA & TECHNOLOGY

• Help establish data standards, privacy and security governance aligned with governance requirements.
• Invest/augment required infrastructure/tech platforms along with capabilities from development to production, to scale agility in conjunction with enterprise data lake for structured/unstructured data access.
• Help establish adaptable architecture/orchestration for enterprise fit-for-purpose LLM models.
• Examine existing vendor portfolio and align with current LLM ops for the selection of vendors.
• Examine/enhance AI models processes and frameworks to deploy/monitor LLMs for the enterprise.

RISK & GOVERNANCE

• Help establish/update firm-wide AI-focused policies and disclosures and risk governance framework for heightened risks from generative language generation content creation; recall all compliance policies; protect IP copyright infringements for potentially diversified content amplification of existing models by self-direction (AI).
• Develop robust testing and monitoring frameworks to measure models’ risks and performance.

PEOPLE & TRAINING

• Train employees on leading practices and business/security risks with use of LLMs.
• Upskill business users with focused training on gen AI applications principles (i.e., prompt generation using developing talent with technical experience, fine-tuning, chunking, prompt chaining, classification, developing enterprise applications).
• Promote use case ideation with product and strategic teams, filtering them through the established intake process.

Summary

Gen AI presents a multitude of possibilities across commercial banks. However, before any investments are made it’s important to evaluate the capabilities needed across key functions as well as overall collaboration in decision-making and top-of-house oversight.

Interesting opportunities such as the Augmented RM Dashboard present substantial benefits to the RM and can elevate both the employee and customer experience in a low-risk way. As emerging nonfinancial companies continue to put pressure on traditional institutions, developing use cases and pilot programs that align with key priorities will help banks realize greater benefits with advancing technologies.

Key highlights include:

• Emerging entrants into financial services have compounded the pressure on existing financial institutions to innovate.
• Gen AI is here to stay. It has solidified itself as one of the most significant technological advancements in decades. It’s critical that banks continue to explore and dive into various opportunities to innovate, with potential risk of being skipped over when customers are looking for the right bank to meet their needs.
• Commercial banks, like most industries, should be primarily asking themselves about where in their organization gen AI should be implemented, and how.
• Three general motivators that result in an implementation continue to be: employee productivity, cost reduction and customer experience.
• Use cases are a necessary first step. There are three distinct areas to think about when assessing potential use cases for a commercial bank: front office, headquarters and back office.
• Before selecting a use case, it is critical to evaluate the client risk potential. Using a decision-making matrix can help with potential use case selection.
• Focusing on the front office could augment RMs’ capabilities and significantly improve customer relationships and processes.
• An interesting front-office gen AI tool to explore is the Augmented RM Dashboard. Using voice recognition, the dashboard works as a personal assistant in the background of a virtual meeting, providing recommendations and next steps to the RM based on conversations with the customer.
• Gen AI should be an enterprise investment, not only tech/AI, ensuring top-of-house oversight.
• Before any investments are made, it’s important to evaluate capabilities needed across areas such as business enablement and strategy, data and technology, risk and governance, people and training.