Insightify
Fake News Detection
The problem Insightify solves
Academic certificates are extremely easy to fake, and traditional verification is slow, manual, and unreliable. Colleges and companies waste time verifying documents, and students face delays when applying for jobs, admissions, or internships.
Insightify solves this by combining AI anomaly detection with Ethereum-based immutable verification.
It automatically checks whether a certificate is forged using AI (layout similarity, seal/logo match, signature analysis) and verifies the certificate’s authenticity by comparing its hash with the secure record stored on the blockchain.
This removes the dependency on manual verification, reduces fraud, speeds up approvals, and provides a transparent trust system for institutions, students, and recruiters.
Challenges we ran into
Challenges I Ran Into:
While building Insightify, our biggest challenge was that both team members were beginners in blockchain and AI development. We had never deployed a smart contract before, and even basic tools like MetaMask, Remix, and Sepolia test ETH were completely new to us.
Some specific hurdles we faced:
MetaMask Setup Issues:
We struggled to switch networks, get Sepolia test ETH, and understand why contract deployment was failing due to insufficient gas.
Remix IDE Confusion:
We had difficulty compiling, deploying, and interacting with our Solidity contract for the first time. Understanding functions like storeReport(), reports(), and how contract state updates work took time.
AI + Blockchain Integration:
Since Insightify uses AI for anomaly detection and Ethereum for immutable report storage, connecting these two parts was initially confusing.
Understanding Gas Fees:
We weren’t sure how gas estimation works or why certain transactions failed even though we thought we had enough balance.
Working With Python Backend:
Creating the folder structure, running Flask, and connecting it to the smart contract was new and required step-by-step learning.
How We Overcame Them:
Followed documentation and troubleshooting guides step-by-step.
Used ChatGPT extensively to debug MetaMask, Remix, and Python errors.
Deployed multiple test contracts to learn how blockchain transactions work.
Simplified the architecture and focused on a minimal working version.
Tracks Applied (1)
Ethereum Track
ETHIndia