Equalis

Arjun Sarao, Bram Ogus, Connor Wilson, Rene Gonzalez Pina

💡 Inspiration

We want to create a system that would help eliminate voter fraud and allow citizens to quickly and efficiently participate in not only their electoral system but their legislative one too. Using the blockchain would help because all voters would be signed with a UUID consisting of the hash of their health card number, passport number, and a special QR Code sent to the voter's address via mail.

📱 What it does

Users sign up by scanning their passport, health card, and a special QR Code sent to them by us. They can then participate in governmental elections as well as in passing legislation. Equalis uses the blockchain for storing election information. All votes are signed with a hash of a voter's health card number, passport number, and a special QR Code sent to the voter's address via mail. The home screen contains upcoming laws as well as a countdown to the next election. On the laws page, users are able to scroll through a list of all laws both passed and not and are presented with an AI-generated summary, followed by important dates, and then the full law itself. The user would then be prompted to either vote for or against the law. When there is not an election going on, the elections page would simply be displaying a countdown, when the election is live, voters can select their preferred candidates.

📃 Full Feature List

Identity verification using OCR and face recognition
AI-generated legislation summary
Election vote counts stored on Blockchain via proof of works
Interactive graphs for vote breakdown for laws and elections

🛠 How we built it

Adobe XD: We designed the mockups in Adobe XD to figure out the UI/UX and the general flow of the app
Flutter: We built the frontend in Flutter using XYZ packages
Python: Python was used for our backend/API as well as for our ML components
- transformers: We used the HuggingFace package, specifically Google's Pegasus model to summarize the legislative documents
- face_recognition: This Python package was used to measure the similarity between a user's passport photo and another photo
- firebase_admin: Firebase SDK for Python
- web3.py: Used to interact with the Solidity smart contract on the blockchain
- py-solc-x: For compiling the Solidity code
- pytesseract: We used pytesseract for the passport verification by reading MRZ's (Machine Readable Zones)
- flask: Used to host our JSON API on the web
Firebase: We also made our backend with Firebase to hold the user data (name, UUID, privilege, votes sent, ect.)
Solidity: Solidity was used to create the smart contract on the blockchain

🛑 Challenges we ran into

Smart Contract development
Making sure the OCR packages were able to run on repl.it
Having Python interact with the blockchain
Figuring out what information to store in a database vs. the blockchain

✅ Accomplishments that we're proud of

Implementing a full smart contract with no previous knowledge of Solidity
Getting it all working, lol
Connecting our app to Rinkeby TestNet

📖 What we learned

Solidity
flask
web3.py, py-solc-x
firebase_admin
pytesseract, face-recognition, transformers

🤔 What's next for Equalis

Refine the user authentication process
Create more robust tooling for creating elections
Add more forms of verification in case people don't have their passports handy
Deploy the smart contract to MainNet or on-premise Private Network
Refine the law summarization

Tech Stack

💻 Lines of Code

389 lines of Python Code
213 lines of Solidity Code
1689 lines of Flutter

🙇‍♂️ Acknowledgements

We would like to thank DeltaHacks for the opportunity to create and develop our idea.

DarkHawk727 / Equalis-API