In our world of digital development, accessibility is vastly improving. However, with the newly started online banking system, we still have plenty of spaces to work on in terms of providing the same convenience to people with special needs.

What it does

Providing banks with data from special-need end-users, such as type(s) of disability, incident severity, spending value and frequency etc, as these are risks factors for big changes in their bank accounts. By analysing the financial information as well as health information, it is possible to ascertain both the degree to which such a person would need support and how quickly they need it.

How we built it

We built a deep learning neural net, with the aim of analyzing the financial information of those affected by accessibility issues and disabilities in finance. Such a model has the capability to uncover the links between various factors within a person life. By focussing on those who may be more at risk, we can specialise and allow customers to be catered to in a more tailored way than has previously been possible. This also allows banks to quickly adapt to changing customer focus, and ensure a more personalised and empathetic banking experience. #dontbeabanker

Challenges we ran into

We found difficulties obtaining legitimate databases to use as data for our neural network. This meant that a database of our own had to be created, using Gaussian distributions for random variables, and uniform and Poisson distributions to generate noise. Unfortunately, there was a misunderstanding in our team, leading to the guy who was doing the neural network to tear his hair out for 8 hours.

Accomplishments that we're proud of

Overcoming all the difficulties we've faced. And we came up with awesome names for our projects. We're proud of the breadth of what our algorithm has achieved as well

What we learned

We've learned how to use Bootstrap and connect a webpage to a domain using We learned to generate a deep learning multi-branch neural network with support for both categorical and continuous variables, and to learn the basics of natural language processing to analyse customer feedback.

What's next for AccessiBank

We're excited to combine both our ability to analyse the needs of the customer through deep learning neural networks and our ability to see in near real time the change of customer opinion by using natural language processing.

Built With

Share this project: