# Solonovel

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [haretha V S](https://cerebralvalley.ai/u/ena_mp4)
- **GitHub:** https://github.com/harethavs07-hue/MasterScan.01
- **Demo video:** https://youtu.be/BCt8NHzaPSc?si=MnhrRc-1sd6rYwDb
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2460

MasterScan(A Software to integrate in Meta glass): A Corporate Watchdog for Your Grocery Aisle
The problem:
Labels say "100% Natural," but the ingredient list often tells a different story. Most shoppers can't decode E-numbers, don't know which additives are restricted in other regions, and have no quick way to tell whether a marketing claim contradicts what's actually inside. When I shop for groceries, I want to know what I'm buying without spending ten minutes searching each product on my phone. Barcode apps give a score, but none of them confront the marketing claim itself.
Our solution:
MasterScan is an AI-powered investigation tool that looks beyond the label. It reads a product's packaging, decodes its ingredients, checks them against regulatory restrictions, and flags where the front-of-pack claims don't match the back-of-pack facts. It also surfaces pricing anomalies compared with similar products.
The Ray-Ban Meta vision
We designed MasterScan hands-free and AR-first. Picture yourself in an aisle wearing Ray-Ban Meta glasses. You say "MasterScan, investigate this," and ingredient warnings and price flags appear in your field of view. No phone, no barcode.
How it works: a 7-agent architecture:
Vision Agent: extracts text from the label image.
Chemical Agent: decodes E-numbers into known identities (e.g., E102 → Tartrazine).
Research Agent: flags international regulatory restrictions.
Finance Agent: detects pricing markups versus comparable products.
Verification Agent: cross-checks marketing claims against actual ingredients.
Pattern Agent: finds formula similarities across a brand's products.
HI Agent: orchestrates the others and compiles the final AR report.
Current status
We built a fully interactive "AR Cyber HUD" prototype in HTML/CSS/JS, inspired by Ray-Ban Meta OS. Users upload two product images, press Scan, and watch the seven agents run an investigation in a neon terminal interface. In this prototype, the agent outputs are simulated so no API credentials are exposed in frontend code. The simulation uses the structured JSON format we designed for Meta Muse Spark 1.3, so the UI is ready to be connected to the real model through a secure backend.
What's next
Connect the Vision and Verification agents to Muse Spark 1.3 via a serverless function, add real regulatory and pricing datasets, and test on Ray-Ban Meta hardware.
MasterScan turns everyday shopping into an informed decision, and turns smart glasses into a real-time consumer watchdog.

## More from Meta Global AI Developer Hackathon

- [Rebound](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2457)
- [Pixel Forge](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2458)
- [Elvis Claros Castro](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2459)
- [MRE Mission Engineer](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2461)
- [me studio](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2462)
- [Jabardast Jetha](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2463)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2460. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt