Top AI Trends in FinTech from Finovate Summit
The following three use cases for AI applications reverberated across the conference:- Hyper-Personalization
- Cognitive Automation
- Risk Management
Adoption of Hyper-Personalization
The origin of app personalization started by targeting specific audience segments or cohorts based on personas. Enterprise apps, for example, worked to differentiate and then individually target small businesses, mid-cap, or large enterprises. From there, consumer apps began to focus on cohorts such as millennials and baby boomers, and then took it a step further by offering personalized products or content based on those segments. The next wave? Hyper-personalization. Utilizing AI and real-time data, hyper-personalization is the next level of app-human engagement, focusing on more individual profile and behavioral targeting to provide more relevancy to each individual. Mobile Commerce (m-commerce) is becoming the poster child of hyper-personalization, offering more intuitive and conversational transactions with end users. The technology has evolved from typing to touching to swiping, with each mode of interaction getting progressively easier for the user. Because of this, voice is the new Touch. After all, what could be more simple or intuitive, more personal than a conversation? Central to the emergence of voice commerce is building connected experiences for customers so they can interact and transact where, when, and how they want. There were a few presentations on Hyper Personalization that caught our eye.- Jason Davies and Katrina Shiu of Flybits gave an interesting presentation on using hyper-personalization to drive engagement with highly targeted offers and outreach. By utilizing data sources like account history, recent transactions, customer profiles, and other out-of-the-box contextual data points, they share how companies like banks are able to engage users at the right time with the best (and most relevant) services and content.
- Datanomers also had a presentation that discussed hyper-personalization. By utilizing hyper-personalization thanks to AI, they accurately predict users’ online interests in real time. This is done thanks to data collection related to user click-streams within a site or keyword analysis, and that data gets turned into hyper-personalized recommendations. The recommended content appears appears, minimizing users from drowning in content, as the information that matters to them most is readily available. The window displays relevant content (including advertisements) and keeps visitors on a website longer while helping to monetize invaluable archives.
From Process Automation to Cognitive Automation
Process automation continues to be a forcing factor in FinTech, especially for banks, to drive down operational costs and improve productivity – and they’re turning to AI-led cognitive automation for help. Use cases range from expediting loan processing using computer vision to delivering faster and better customer experience through virtual assistants. One interesting company making strides in this direction is Finsend. Dropping by their booth at Finovate Summit, I watched their demo of their AI-driven Banking Dispute Platform (BDP) which was designed by financial industry veterans who have first-hand professional understanding of the unique challenges around dispute resolution. The core of their solution is an AI-driven engine that streamlines the entire dispute process by offering real-time reporting, batch processing, fraud monitoring, and an enterprise user account area. By deploying BDP, Finsend claimed that banks were reducing the costs of chargebacks by an estimated 25 percent and increase cardholder satisfaction – showcasing how cognitive automation can help drive down operational costs while improving productivity (and an added bonus of customer satisfaction).Risk Management and AI
“With great power, comes greater responsibility.” — Uncle Ben to Peter ParkerAn often-overlooked fact is that financial institutions absolutely must have a security-first approach due to an interesting matrix of an ever-increasing digital presence thanks to customer demand, combined with highly personalized customer data, a risk-heavy environment that requires proper security protocols. Skipping the traditional thread around PII or GDPR here, and instead focusing on an interesting thread of how protecting products and engineers against malicious actors can be done through the use of AI. It’s easy to endure sleepless nights thinking about fraud prevention, denial of service, identity, and data theft for customers, who trust an AI platform with their data. But that AI platform can actually help with risk management. At Appen, we already scan user traffic for malicious activity using AI-based pattern matching of packet data traversing our cloud network. That’s not the only approach to risk-management with AI. A keynote from Illuma Labs at Finovate Summit examined both security and automation. The team from this Texas-based startup talked about Illuma Shield, their real-time voice authentication service for call centers. Milind Borkar and Manuel Gean discussed how the convenience of automated banking has sacrificed the personal connection between long-time customers and their banking institutions, who bombard them with requests for identification at every contact. Their tool passively analyzes a caller’s voice over natural conversation, resulting in higher authentication accuracy in a fraction of the time and enhanced security against fraudsters. Some of our customers train their voice biometric algorithms with the data we help them annotate.