2020 was a different year. It helped us understand the value of social living and society more broadly. The #BlackLivesMatter was not just a movement but a realization of how the world has still not matured even decades after the civil rights movement. The discrimination is real and it happens with everybody in minority. Whether you are a single mom, a hispanic dreamer, an African American entrepreneur or an Asian engineer, the world sees and treats one differently. Last year was a wake-up call for the society to mature and this project is just trying to eliminate the bias in the lending industry because access to capital should not be based on the bedrock of discrimination.
What it does
Guppy is an AI platform that matches loan seekers with guarantors. The AI analyzes an individual's profile based on the credit information and searches for guarantors that have a credit score that can boost the loan seeker's chances of getting an approved loan. Unlike traditional loan guarantee applications, Guppy provides a % of the APR to the guarantor so that more and more guarantors join the platform and develop a steady-state cash flow by being a guarantor to the loan seekers who otherwise wouldn't have gotten a confirmed loan. At all points of time, Guppy keeps all the profiles of loan seekers and guarantors completely anonymous to avoid any human bias from jeopardising an application.
Loan seekers get a guaranteed loan. Banks get a guarantor wherever they don't feel the loan seeker's credit score matches the loan requirements. Guarantors get paid to be a guarantor on a loan seeker's application.
How I built it
- MEAN stack
- FINASTRA API
- FB Login
- Google Slides
- Adobe Photoshop
- Loom video recording
Challenges I ran into
Mostly tech challenges since I am a single-person team working on coding, designing, ideation, video recording and everything else. But beyond the hackathon all of these are fixable.
Accomplishments that I'm proud of
Idea. I think the idea has a huge merit and can revolutionize the multi-billion lending market.
What I learned
The alarming state of discrimination in the lending market.
What's next for Guppy
The code is still very scrappy and the model is untrained. The model will need to be in the dry run for much longer to have a promising classifier precision and recall.