Discover how Generative Artificial Intelligence (AI) is not just a buzzword but a game-changer in the financial industry.

In this article, I unveil actual demos that showcase the transformative power of Generative AI in enterprise solutions in WealthTech.

The genie is out of the bottle with Generative AI. Enterprise service providers are in no shortage, but who and how exactly are businesses walking the Generative AI talk, is not clear. We all need to see more demos and use cases.

Surveys from consultants and AI tool providers mask the actual usage under the umbrella term AI.

GenAI is only a very small part of the Machine Learning (ML) tool set, and often automation and business intelligence tools are lumped into the 'we use AI in our businesses'.

There are plenty of examples in financial services to show how AI is deployed extensively, but it has been low-level automation, predictive analytics, some Machine learning, and GenAI are just emerging.

Rules-based chatbots in banking have been around for a while, rules-based workflow automation tools for report generation are frequently marketed as AI-driven solutions, and a large part of the huge variety of traditional fraud detection systems in finance use a set of hardcoded rules to flag transactions as potentially fraudulent.

ML deployment has also been embraced in financial services, with my favourite use case in lending. FinTechs and incumbents have been increasingly analysing traditional and alternative data sets using machine learning algorithms to determine eligibility for credit. But this, again, is not a Generative AI use case.

Generative AI could be used in lending to enable a more interactive, collaborative underwriting process. Imagine a borrower interacting with a lender to determine the next best action that will increase the borrower's ability to qualify for a loan. This would be a totally new, personalised customer experience and a potentially new revenue stream for lenders who can offer such a scalable digital solution. ^[1]

I have personally seen three demos of enterprise solutions deploying Generative AI. All three cases do not use client data but market and research data. That does not mean that the data sets used for training are public data. They are information that clients pay to access, and essentially, Generative AI is changing completely the paying customer experience. Generative AI is not only speeding up access, but also improving the experience as it is in natural language, and it provides new ways to interact with the data. Discovery is clearly different. Less experienced users get a quick upgrade, and expert users can focus on better insights, decision making and problem-solving in their business.

I have written about how Moody’s and Morningstar are using Generative AI trained on their extensive knowledge libraries. ^[2]

On Tuesday, I attended the successful annual Swiss WealthTech event organised by The Wealth Mosaic at the SIX convention centre.

Roy Kirby, head of core products at SIX, shared with us three demos of how SIX has deployed Generative AI for their clients (banks, asset managers, etc) who use their market data terminal and their trade surveillance tool.

The SIX market data terminal use case
SIX has built a ChatGPT-type interface into their market data terminal. This means that users don’t have to remember the keystrokes to obtain the data they need. During the demo, we saw a comparison of an expert user of the terminal searching for a specific set of information around the top 5 gainers in the Swiss stock market in comparison to using the GenAI-powered interface. Expert user of the terminal needed 46 secs, and Generative AI took only 10 seconds.

That means that all level users will now take 10 seconds to complete this search. They can all afford to go deeper or broader in their search, as it has become so much easier to interact with the data.

Generative AI will also tell the user the source of the data.

SIX has also deployed Generative AI capabilities to deal with the flow of requests around adding a new instrument or fixing information on an existing instrument on the market terminal. These are typically communicated via email, and now Generative AI deals with all these requests with no human intervention (24/7 and with no burnout).

SIX reports that Generative AI managed to resolve over 4000 requests within seconds with no human intervention needed and with great customer satisfaction. This also means that customer service agents can deal empathetically with more complex customer problems. SIX customer satisfaction metrics also went up.

The SIX trade surveillance tool SICAM
The demo (in beta mode) showed us how users of this tool (from banks to brokers) are empowered when the SICAM tools alert them about suspicious trading activity. The compliance officer`s investigation is an elaborate process to determine whether this is a false alert or not. In all cases, the officer has to justify his/her decision.

GenAI is being used to assist the compliance officer in the justification of a false alert. GenAI collects all relevant information on behalf of the compliance officer to support a false alert decision. The decision remains with the compliance officer, but GenAI has not only saved they at least 10 minutes per case but also the quality of the reporting is hugely improved. Currently, the workload for a compliance officer is increasing, and their reporting is typically a cut and paste from the SICAM tool with some part of the justification that does not include all relevant information.

Conclusion
As we’ve seen through these vivid demos, Generative AI is redefining the very nature of B2B customer interaction and compliance. The potential is huge, and the journey has just begun.

Financial services providers need to tap into data science expertise through collaboration and cross-industry pollination to design use cases that have an impact on their businesses.

[1] BankThink — AI could make underwriting a collaboration between bank and borrower

[2] A Moody's & Microsoft GenAI collaboration for the Era of Exponential Risk

Mo, the First Personified Next-Gen Chatbot for Investment Research

Read the original article here.