BlockImage
Privacy Imaging with GPT (Giving Privacy Toast)
Created on 30th November 2025
•
BlockImage
Privacy Imaging with GPT (Giving Privacy Toast)
The problem BlockImage solves
People upload files online every day (especially in GPT's), but it’s hard to know if those files are safe, private, or tamper-free. Once a file is uploaded, anyone running the server can see it, change it, or lose it without you ever knowing.
AI Vault solves this by making file uploads private, secure, and verifiable.
Challenges I ran into
While building this project, I faced a few issues:
File encryption errors – Sometimes the browser wouldn’t properly convert files into encrypted data. I fixed this by switching to a more stable hashing and encryption flow.
Backend connection problems – The backend and AI worker kept refusing connections. I solved it by running both services on the right ports and adding proper error handling.
QR code generation bugs – At first, the QR code wasn’t showing in the frontend. The issue was just a wrong Base64 string format, which I corrected.
Storing data on Aptos – Writing transactions and making sure they worked on testnet took time. I fixed it by testing with smaller payloads and checking each step on the Aptos Explorer.
UI breaking due to imports – Some React components failed because of wrong imports or missing icons. Cleaning and reorganizing the imports fixed the issue.
Overall, each challenge pushed me to debug smarter, test small steps first, and keep refining until everything worked smoothly.
Tracks Applied (1)
Best Use of GitHub
GitHub
Cheer Project
Cheering for a project means supporting a project you like with as little as 0.0025 ETH. Right now, you can Cheer using ETH on Arbitrum, Optimism and Base.