Inspiration
I was studying content parsing in js and got an idea of creating a tool which can detect for specific patterns on a website, and this idea eventually lead to NarcoScan.
What it does
It scans website through provided url and uses html parsing, scraping and llama ai model to detect if the given website contains or sells or promotes illicit drugs.
How we built it
Html for frontend, Tailwind css for styling , node.js for backend, and chart.js for bar charts. Initially it was just keywords scanner but then i integrated it with groq llama 3.3 70b model so that it now detects patterns and summarizes result.
Challenges we ran into
The only challenge i faced was creating it in limited time.
Accomplishments that we're proud of
I am proud of bringing this idea to life in such short period of time and also learned many new things.
What we learned
In the process i learned working in tight schedule and also much about ai models integration.
What's next for Narco Scan
Next, I have plan to modify it and make it a live drug content monitoring system which uses web crawler, scrapy, and ai models to monitor all such active websites and provide all illegal drug related activities happening on the internet in real time. It could help narcotics department and other govt. agencies a lot.
Built With
- chart.js
- express.js
- groq
- html
- javascript
- node.js
- tailwindcss
Log in or sign up for Devpost to join the conversation.