Tedious process of getting a loan online requires providing a lot of information which our banks already have. PSD2 kind of APIs can ease the process much giving a possibility to automate it. After filling all of the required fields it takes time for the Bank to judge what is the customer's current situation and decide if he can get the loan. This process includes validation of the provided information, check with 3rd party vendors about credit history and such. In case when the loan submission will be rejected, the customer is put into situation where he must repeat the same process once again from the very beginning but at other bank or online store.
What it does
Open's new opportunities for cooperation between Banks and stores. Introduces the SPI which can be implemented by a store to provide recommendations. Those recommendations are fetched with some criteria to find alternative products for which a customer can take a loan. AI engine matches recommendations with customer expectations defined by the initial product he wanted to buy. Together with rejection the bank can respond with new pre-processed options of loans or alternative products with a lower price. By doing this the store can keep the customer and the bank still have a chance to sell a loan. Offered alternative products are already pre-processed so it is guaranteed that a loan will be given what improves customer experience who isn't required to repeat the whole process one again.
How we built it
Our solution consist of few elements.
Store which offers some goods
Front-end in Angular and back-end in Java (Spring Boot). The store exposes an API which is complient with a SPI we defined. Thanks to this SPI other vendors could integrate very easily. The SPI defines a way to fetch recommended products when the customer is not able to complete a process of getting a loan successfully.
Bank which offers loans online
We have developed a sample bank application with a front-end written in Angular and back-end in Java based on SpringBoot framework. The app allows a customer to get a loan online. Filling a form has been automated with a usage of APIs exposed in FFDC. Summary page, displayed to the user after the loan submission rejection, contains a list of pre-processed alternative products which user can buy, taking a loan, with a single mouse click.
3rd prat bank which authorize the customer
A front-end application which represents another bank where our customer has an account and from which we can fetch a customer profile. By doing this our application automates the process of gathering information from the customer, who doesn't need to provide them manually anymore. The only thing he needs to do is to login to his bank, and consent the access to his sensitive data. To implement that we've used Account Information (US) and Customer Profile APIs which are exposed by FusionFabric.cloud platform.
Challenges we ran into
Access to customer's sensitive data requires his consent. PSD2 APIs exposed by FFDC have implemented a consenting mechanism, but due to the lack of some privileges it was easier to use US alternatives.
Accomplishments that we're proud of
We have delivered running software, with decent visible business value, giving understandable presentation, starting the hackathon without a clear idea of what we gonna do, enjoying the event!
What we learned
With a tight time constraint we had to cooperate efficiently to deliver that much stuff. This nicely integrates a team of people who are technology enthusiasts.