It’s conference season, and DigFin has sat in on a slew of fintech demos. Here we highlight five companies using artificial intelligence to build digital assistants or chatbots for use by financial institutions, based on presentations at Finovate (a fintech demo event) and broader industry conferences. They share a common theme of using tech to help people working in finance to add more value – in sales and client service – by cutting out manual work. (See also our stories on how J.P. Morgan is using a Hong Kong chatbot maker, and meet the company behind China’s most successful banking chatbots.)
AlphaBot is designed to help people in the alternative-investment space to efficiently collect data on products, including crypto assets; and to help hedge fund managers sell to investors or improve their client-relation services.
The company behind AlphaBot is U.S.-based AlphaTech Investment Solutions. It argues that alternative fund managers find it hard to source and collate the data required for analysis of their portfolio models and construction. Once that work is accomplished, hedge funds then have only static PDFs as a means of presenting their arguments to their institutional investors.
AlphaBot presents itself as a network providing APIs to allow hedge funds and other alternative managers to pull data on asset classes from vendors such as BarclayHedge and Morningstar, as well as crypto-currency feeds. They can use the AlphaBot platform to visually compare and tweak data sets, benchmark results, generate reports, and present portfolio ideas.
This Hong Kong- and London-based chatbot company uses A.I. to help insurance companies sell online by automating underwriting. It’s designed to let them reach customers directly, rather than through agents or other intermediaries.
It’s chatbot advisor, “Anna”, selects suitable products for potential customers based on a few metrics, such as type of insurance policy, cover amount, gender, date of birth, annual income, and smoking habits.
This takes place via a user-friendly experience that is more akin to a chatroom conversation than a questionnaire. Anna will ask the customer their name, and then go into underwriting-related questions like your occupation and whether you, say, scuba dive or participate in other risky activities…and on through more items around health conditions and the like.
Users then input their government-issued I.D. details, contact information, and agree to the insurer’s T&Cs. Then Inzurer generates a policy and asks the user when it’s O.K. for the insurer’s human agent to get in touch. The entire sales process is designed to take 10 minutes, including for life and health insurance policies.
This India-based company‘s intelligent banking system combines voice and text to allow banks to communicate with customers. During the demo, a Payjo executive called the chatbot and demonstrated its ability to shift between languages, although the bot paused before every response; its voice was also robotic.
The chatbot’s scripts are designed to hook and upsell clients. For example, when asked if a credit card charges an annual fee, it knows to suggest alternatives in case the answer is “yes”.
The chatbot’s programming is embedded in the sign-up forms for products such as a loan or a card. It has been designed to help potential customers navigate a bank’s online paperwork, as a human person would do at a bank’s branch. Customers can ask it to clarify anything confusing about filling in a form. And it switches easily between text and verbal prompts. It can also be programmed to send a push message or phone a customer who drops out of the sales process mid-stream.
Payjo’s execs say it is about to be adopted by a large bank.
This company’s chatbot is designed to offer mass personalization to a bank’s customers, by making conversation sound natural. During the demo, although its response was a little slow, the robot did sound much more like it was a person having a normal conversation.
Israel-based Voca.ai designed its chatbot to help with lead generation, lead qualification, upselling, and scheduling appointments.
The technology involves a deep-learning process of feeding its call script with hundreds of thousands of human collection calls. The machine is a layer of call scripts, with trigger phrases and contexts designed to identify customers most willing to pay.
Users would have to similarly train the machine to handle their specific services. Voca.ai also comes with a visual dashboard so a bank or card issuer can predict how the system will respond, as well as to connect it to their external systems.
This India-based company has developed a personal digital assistant specifically for mobile use by sales reps, and says it now has 75,000 salespeople using it.
The app gives salespeople recommendations, suggested actions, and helps them log data back into a CRM like Salesforce.com. It integrates into a salesperson’s office calendar, suggests venues where people can meet prospective clients, identifies prospects who are in a salesperson’s vicinity, and keeps a file of the salesperson’s calls and meetings, as well as a bio or snapshot of customers and prospects, including tips on what products they might buy.
It also provides them with a dashboard showing their progress, a history of their interactions with each client, and a score of deals or fails. It monitors the length of calls, asks salespeople to rate the quality of each call, and let them add notes. This information feeds back to their managers, in real time.