Dashboard of MoneyTelly
All the experts agree that voice interaction will be one of the next big things. We are convinced that the reason for that is the simplicity of the interaction. People are tired of using hundreds of different apps for small tasks, they very much prefer to have one touch point. For the same reason, many young people are not really excited about their banks, especially not outside of the Fintech area. For a traditional bank like Raiffeisenbank, it is immensely important to keep in touch with the young generation, and we have the strong believe that a integrated, natural touch point like MoneyTelly will be the answer to the challenge.
What it does
MoneyTelly provides a natural language, voice enabled interface for the most occurring tasks in banking: Checking the balances, doing transactions, etc.
Additionally, we integrate with one of the most popular finance-related apps on the market. Splitwise helps you share bills and keep track of expenses in a group of friends or in a family. You can for example pay the bill in a restaurant, and use MoneyTelly to split the bill exactly like you want to. It also keeps track of your balance with your friends and obviously, you can pay your debts with just one voice command.
When it comes to banking, security plays a big role. We looked into recognizing a user via voice, but the state of the art doesn't look secure enough. However, we noticed that with contact-less credit cards, you don't have to enter a PIN for smaller payments. This confirms our impression, that convenience is one of the most important features nowadays. Recognizing the voice pattern can give enough evidence for transactions up to a certain amount, after that another security measure has to be taken.
MoneyTelly doesn't stop there. We also implemented the functionality to access data collected by smart home devices like a connected heating. Used in conjunction with historical data, not only regarding energy consumption but also your income and expenses, we can predict your balance.
When we see, that the balance is constantly increasing, we want to offer the user investment plans. At this point, a traditional bank like Raiffeisenbank can show its strenghts. People trust it, and they know they have a life-long partner at eye-level. Another integration of MoneyTelly, this time with the calendar, makes scheduling an appointment as easy as possible. The user comes with a good feeling and can leave with an investment plan, a win-win situation for both him and the bank.
How we built it
We have created an Alexa Skill that acts as the voice interface for the application. The frontend is a node.js web application served by an express server in the backend. There, we can connect via API's to all sorts of knowledge sources, including the CRM system of the bank to enable a personalized experience. Alexa communicates over the Lambda function hosted on the aws backend via Websockets with our Frontend.
Challenges we ran into
Alexa was created to act as a voice interface. We believe, that for a banking application, it is important to also give visible feedback and information. Since Alexa wasn't created for that, it was rather difficult to connect Alexa with a web frontend.
Accomplishments that we're proud of
Overcoming the challenges and providing a clean architectures including WebSockets to give the user a seamless experience - both with his ears and his eyes.
What we learned
While testing, we realized over and over again, how important convenience and a natural way of interacting with a service is. Keeping this in mind while iterating over this idea can really make the difference compared to the competitors on the market.
What's next for MoneyTelly
When it comes to investment, people tend to be very conservative and are rather 'losing' money due to inflation instead investing it. We believe, this is mainly caused by a lack of knowledge and by unpersonalized offers. MoneyTelly can learn over time, what type of human the user is by looking at his expenses. The way someone talks to a machine already tells us a lot about his personality, and by asking the right question, MoneyTelly can get to know the user very well, which in the end will lead to more satisfied users and more investment plans for the bank. There are more integrations to implement.