-
mobile_authorisation_successful
-
mobile_authorisation
-
mobile_landing
-
web_authroisation
-
web_team
-
web_landing
Inspiration
With most of the members of this team coming from overseas, they’ve had their fair share of difficulties dealing with the bank. For example, one cannot apply for credit card easily despite the long standing good credit history in home country. Under globalisation, people studying or working aboard or emigration become commonplace. Having to build up your reputation in terms of credit score again from the beginning when you’re only here for maybe a few years seems inconvenient. That’s the main inspiration for this project, to alleviate some of the pressure that international students face when going abroad. We believe it is also beneficial to banks at lower cost to access more extensive credit data of potential borrower. The banks have a fuller picture to the borrower status for making informed decision in granting credit, and potentially reaching the market with previously unknown information.
How we built our project
We dived into our project with some discussions and brainstorming first. Since our topic was developed relatively late, we decided that it’ll be best to not be entirely focused on coding the small details, but rather to come up with an outline on how exactly we are going to implement this. Our project was then split into two parts. One as a prototype, displaying the mobile app interfaces to show the full process of our credit score management. The other part is a web page, working out draft codes to implement the functions.
What we have successfully built
Putting useful data that UK credit reference agents or any lending companies value, into use is certainly a crucial part of building a trustworthy, adaptable credit score calculation system. So the first thing we built is a data collection page, to collect personal information such as credit history, negative financial events etc. We created app interfaces for the manual uploading of credit history and other additional credit details with a requirement of supporting documents to the specific event. The authorization interface allows an alternative way of inputting personal details through access to previous financial agencies. Users’ banking history or even past credit reports are supposed to be imported after requesting access to it. Having all the records required, our mobile app will calculate and display a predicted credit score. We have also created interfaces for users to record their own credits and loans applications. So that CreditGo can remind users of loan repayments to not affect their credit scores.
On the other side, we have interfaces designed for lending companies. Companies, after registering an account, can gain access to the user's predicted credit score and past financial activities through the authorization of the users. Loan lenders and credit agencies can reference such information to make decisions on giving credits.
Challenges we faced
Technically, most of us have no prior experiences on website building. We have no idea how html works and hence can only rely on prototyping tools such as Figma to present concepts we have in mind. As fresh new computer science students, we also have a limited knowledge on designing app interfaces and everything about user experiences. Time constraints has also limited the product we can create (and the time we can sleep). Having to have made up our minds to work on this specific topic after a solid 5 hours, had us working on the project overnight. Frantically finishing off the websites at 11:59 has of course affected the completeness and quality of our hack.
Practically, we are aware of different regulations on sensitive data collection and cross-border data sharing. It is not sure for now if our hack would actually work out without tons of research on laws and legal details. Collection of data from financial organisations also requires collaboration with those companies which might not be possible without an already-built-up reputation and credentials. Security issues such as the prevention of data breach, given a lot of sensitive information is being implemented is also something we will have to, but have not yet considered in our project.
What ideas we have in mind but have not yet built
We planned to make use of AI for the calculation (or re-calculation) of credit scores with existing credit history from aboard. A machine learning model can also help to compare our predicted credit score with credit scores from UK credit referencing agencies. And by doing so, keep checking and improving our predictions accuracy. However, we currently have no access to such data (and time, of course) to train a new model which could potentially be implemented.
Given the nature of the data we handle, it is understood that extra care is necessary in terms of security. The data server has yet to be built. It is also believed that cloud server may not be the best solution as transmitting such data via cloud server might mean that the data is transmitted to a third party which may lead to data leak.
Also, we have not built a bank user interface for importing and exporting credit data to out platform. It will be raw data as we know they have their own algorithm for calculating credit score. SME may also use the platform to build the credit score and upload different information.
What we learnt
Sleep is for the weak!!!
An important lesson for us is that working in a team with a clear division of labour according to one’s expertise would smoothen the whole process. The use of tools such as GitHub and Figma on a collaborative basis are largely new to our team. These new implementations have proven their merits in a collaborative project.
How do we want to further develop/ make use of this in the future
As we are all freshers to the world of computer science, we were only able to work on mockups using interactive prototyping tools such as Figma to work on the design and workflow of the system/application. A website is created on github, but it still lacks a functional backends. In the future, we may build the application with Flutter and integrate with support from Google cloud and local servers. The use of Flutter enables cross-platform development for our app, as it is expected that users will be coming from different platforms. We also hope that in the future the application could indeed be materialised, for the good of new immigrants.
Log in or sign up for Devpost to join the conversation.