About Us
About Us
First of all, Bank of Tomorrow is
not a real bank. Rather, this
website was set up to make people aware of the
algorithms underlying banking decisions, such as
mortgage applications.
AI has been affecting many industries, including
banking. Here, we have created a fairly simple
Machine Learning (ML) algorithm which will approve
or reject a mortgage application. The idea of this
project is to let people interact freely with the
algorithm, to better understand how they work, and
essentially why applications may be rejected. For
more information on our algorithm, please continue
reading below.
Our Algorithm
Our Algorithm
The fundamental goal of our algorithm is to detect
mortgages which are at risk of defaulting. Mortgage
default is bad news for any bank, as it involves an
often lengthy and expensive legal process, where
overall money is lost. This algorithm was created
with the idea of rejecting potentially 'risky'
mortgage requests, thus minimizing costs for the
bank.
In this algorithm, we have worked with a 2017-2018
dataset provided by
Freddie Mac
in the US. It is one of the few publicly available
datasets about mortgage default. You may
(reasonably) ask the question: "How relevant is an
algorithm that was created from a 2017-2018 US
study, if I want to try to get a mortgage in the UK
in 2025?". Every ML model has its limitations, and
ours is no exception. In fact, our algorithm was not
designed to give you a whole lot of actionable
insight. Our idea was to empower consumers, by
letting them interact with an ML algorithm that is
similar to those used by real banks. This way,
consumers can gain insights into how ML banking
models work under the hood, and understand the
hidden biases in these algorithms. If this interests
you, please fill in the form below to start
interacting with our algorithm!
