Understanding the four AI streams in banking. Which one is right for you?

November 3, 2018

Huge shifts are occurring in banking because of technology. The potential of AI to help banks perform better is getting a lot of mindshare from Banking CXO’s. We have identified the top 4 streams that banks are currently using AI in. Along with your strategic direction and strengths, this will help you narrow down where to focus your AI initiatives.

Chatbots and Virtual Private Assistants

Front office is seeing the most mature AI cases in banking. Chatbots and Virtual Private Assistants interact with customers to solve problems before human service representatives get involved. They are easy to make and run, cheaper because they replace expensive humans, and customers are comfortable with them. This makes them the low hanging fruits in an AI strategy. With 88% of Millennials preferring digital customer care channels, 24/7 chatbots are perfect for them. Banks can also use them for technology service desk requests internally!

Chatbots rely on natural language processing and specific use cases defined by their development team. Advanced chatbots, relying on predictive analytics and working with clean high-quality data, can go further and become Virtual Private Assistants. By tracking spending habits and access to financial history, they can help consumers manage their budgets efficiently. For example, Erica from Bank of America not only manages daily transactions but also gives smart personalized recommendations.

Operational efficiencies

Operational efficiencies using AI is a multi-hundred-billion-dollar opportunity for banks. Leading the charge here is Robotic Process Automation. Using technology like screen scraping and AI these robots are easily used in legacy scenarios and can automate workflows. The ROI is quick and RPA technology mature, allowing for certainty that business benefits are achieved. The 234-year-old Bank of NY Mellon Corp is a great example of the benefits of RPA. They implemented over 220 bots to streamline and automate processes. Annual savings were over 300,000 USD. Additionally, key metrics showed significant improvement: examples are 88% improvement in processing time, 66% improvement in trade entry turnaround time, 100% accuracy in account closure validations, .25 second reconciliation of a failed trade compared to 5-10 minutes by a human and so on.

AI specifically used for annual commercial agreements has also been a resounding success. JP Morgan used to spend 360,000 hours to perform annual manual review of 12,000 commercial credit agreements. Using machine learning, they built a platform that can analyze and pull out important points for consideration in a few seconds!

Customer profiling

Before the advent of disruptive technologies, master data management was considered key for banks. The more complete, clean, consolidated, and detailed view you had of the customer, the more you could cross-sell and upsell to them. Predictive analytics used on such data can reap even richer rewards. By building full psychographic profiles of customers using not just financial data but also social, commercial, and personal data, often via usage of mobile and applications on it, banks can create hyper-personalized offers for customers. They can time intervention to pitch those products to maximize conversion.

Tencent, Ant Pay alliance and JP Morgan have been using this approach to accelerate their business. Banks have also been using predictive analytics and psychographics for augmenting their credit underwriting and risk underwriting capabilities.

Fraud and money laundering

Using pattern recognition technology, AI can point out anomalies with serious implications for the fraud and money laundering capabilities of an organization. The key here is to extend AI to unstructured data as well as structured data for the richest dataset mining possible. From finding suspicious trading patterns to using audio and video recognition to identify fraudsters already exploiting the system, this is a use case that has a lot of banks seriously investing to protect their money and brand. The UK government is on track to recover £7 billion in additional tax revenues because of its ability to identify this kind of fraud.

Conclusion

Banks have a lot of work to do to find out how they can best employ AI tools and technologies for benefit but the advantages of being a leader are becoming clearer daily as early adopters start showing results that improve the top and bottom line.

 

Leave a Reply

Your email address will not be published. Required fields are marked *