DAuto

Github Link: 

https://github.com/VictorChan880/dAuto

Inspiration

The inspiration behind our team’s creation of d⋅Auto stemmed from a desire to revolutionize the insurance industry by addressing its most pressing issues: inefficiency, lack of transparency, and vulnerability to fraud. We envisioned a peer-to-peer (P2P) blockchain system that leverages real-time claims processing, ensuring swift and accurate settlements. The security inherent in blockchain technology, coupled with the transparency of transactions, aims to build trust between our community of users. By integrating machine learning, we sought to enhance fraud detection and streamline operations, making the entire process more efficient and reliable. Our goal was to create an insurance model that is not only secure and transparent but also innovative and customer-centric, ultimately transforming the way auto insurance is perceived and managed.

What it does

Our blockchain auto insurance solution is designed to streamline and enhance the entire insurance process through cutting-edge technology. Utilizing a peer-to-peer (P2P) blockchain framework, it facilitates real-time claims processing, enabling swift and accurate settlements without the need for intermediaries. The system ensures security and transparency by recording all transactions on an immutable ledger, providing clear and accessible records for both insurers and policyholders. Additionally, our solution incorporates machine learning algorithms to detect and prevent fraudulent claims, further safeguarding the integrity of the insurance process. By leveraging these advanced technologies, our platform offers a more efficient, secure, and trustworthy auto insurance experience, transforming how claims are handled and policies are managed.

How we built it:

Backend: Built Smart Contracts on Near Protocol and built ML model on Tensorflow and Flask.

Frontend: Developed using JavaScript, HTML, & CSS. We utilised the NEAR Protocol Package, incorporated various elements to enhance usability, and ensured all necessary information was accessible. Additionally, we added icons and designed a logo using Canva.

Challenges we ran into

We encountered several challenges, particularly in integrating our front-end and back-end systems. Initially, our front-end was built using TypeScript, which provided strong typing and good code quality. However, our blockchain was developed in Near, leading to compatibility issues between the two environments. This discrepancy required us to rethink our approach to ensure seamless interaction between the front-end and back-end components. To address this, we decided to create a new user experience (UX) that harmonized with our technological stack using Javascript, HTML, & CSS. This involved designing a new website from scratch, utilizing and building upon the NEAR Protocol Package. The process was challenging but ultimately resulted in a more cohesive and efficient system that met our project's goal.

What we learned

Through this project, we learned that it's okay to restart or stray off the initial plan when things aren't working out. This taught us the valuable lesson that sometimes starting over is the best path forward. 

What’s next?

We aim to continually improve our machine learning algorithms for fraud detection, incorporating more advanced techniques and larger datasets to enhance accuracy and reliability.