Dream 2eam iNTUition

Scam Prevention Visualization targeted at Youths.
Correlation Heatmap of a fake customer dataset we have used for this hackathon.
Decision tree created to identify customers who are more likely to be scammed.
Scatterplot matrix showing the data we used for this hackathon.

Inspiration

We were scammed in 2016

What it does

Scam prevention and detection while educating the masses Please check the GitHub repo and view our Jupyter notebooks for actual demo

How we built it

Jupyter, Figma, and lots of effort researching possible solutions and evaluating how we could bring added value to our target, Goldman Sachs. Solutions should feel related to the business domain and touch actual real cases instead of being fully focused only on the technical portions and forgetting about the practicality. A 99% scam detection system would be useless if it could only be utilized by 1% of its actual user base.

Challenges we ran into

Limited datasets due to secrecy of customer information. This was reasonable given how banks handle their trust with customers and its importance. Attempted to recreate customer data by populating with false information but limited resources, hence referenced to other customer data datasets and utilized already existing variables.

Accomplishments that we're proud of

The video looks aesthetics and compacts our research and development into the 4 minute mark. We iterated our work process through agile work cycles with scrum methodology in mind to hit as much points as we could possibly do within this hackathon duration. We picked up new knowledge and shared thoughts with each other, revising what should be done and how our solution would be relevant instead of creating something just for the sake of showcasing tech.

What we learned

There are many scams out there and how these scams are able to successfully get their objectives met, either through gullible targets or lack of infrastructure to detect these scams early on. There is always a balance between prevention and detection as one acts before something has happen and one only when it has already happen. Remedy is relatively easier compared to prevention as we are never able to 100% predict the likelihood of something happening.

What's next for Dream 2eam iNTUition

Doing well for our tests on Monday! SC1007 is really challenging too. We hope that our solution would be able to bring in new insights and shed light on possible adoptions that Goldman Sachs can use.