Inspiration
Scam emails are a massive issue that affects millions of people worldwide, a web extension that would help people identify scams would be incredibly useful.
What it does
It appears as a box where you're able to select highlighted text for it to scan. It will then display to you whether or not something is (most likely) a scam and if it is a scam, provide the reasoning for it.
How we built it
We built this project in Rust, where we made use of the Jev AI API, and various OpenAI compatible APIs all over OpenRouter. The text would be fed into Jev to verify whether it is a scam, then fed into multiple decision models to create a human readable output which also aggregate the result. We also used Dioxus for the front-end
Challenges we ran into
We suffered communication issues that impacted our progress early on.
Accomplishments that we're proud of
We adapted Telluvian's advice and ideas to expand it to include AI models acting in parallel
What we learned
We learned how to use AI APIs via OpenRouter in Rust and elements of front-end development and deployment
What's next for Scamvoidance
Heightened multimodality: Voice-to-text, images and advanced auto-detection features.
Built With
- css
- dioxus
- html
- javascript
- jev
- openrouter
- rust
- webassembly
Log in or sign up for Devpost to join the conversation.