# KAGUA-LIVESTOCK

- **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:** [Jean claude adjanohoun](https://cerebralvalley.ai/u/jeanclaude), [Soumyajit Das](https://cerebralvalley.ai/u/Soumo17), [Bonface Alfonce](https://cerebralvalley.ai/u/bonface_alfonce)
- **GitHub:** https://github.com/BonfaceAlfonce/Meta-Hackathon---Kagua-Livestock
- **Demo video:** https://youtu.be/RMdK7oU7mJ8?si=XGq4QHp4qgla0ZPL
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/683

Kagua Livestock is an AI-powered smart environmental monitoring and early warning platform designed to protect smallholder poultry farmers from devastating flock losses caused by unmonitored heat stress, toxic ammonia spikes, and disease outbreaks.

Traditional livestock monitoring requires constant physical presence, meaning critical hazards like high temperatures or poor ventilation often go unnoticed until half the flock is affected. We built Kagua to bridge the physical farm to the cloud in real time:

Sense & Ingest: Low-cost ESP32 microcontrollers with sensors deployed directly on-site in zones like the Brooder house and Feeder house stream continuous hourly telemetry (temperature, humidity, ammonia, and CO2) alongside automated camera snapshots into a secure Supabase backend.

Reason & Understand: The platform leverages Meta’s advanced AI models—using SAM 3.1 to segment flock behavior and Muse Spark to analyze sensor trends and images. Furthermore, our Outbreak Shield, and farm advisory agents run automated risk assessments and use Muse Spark to generate plain-language explanations and farmer warnings when unusual patterns or contamination signs appear.

Alert & Protect: Powered by custom n8n workflow automations, the system instantly triggers critical real-time alerts straight to the farmer's WhatsApp when environmental thresholds are breached, ensuring rapid intervention regardless of distance.

Built with React, Vite, Muse Code, Supabase, n8n, and the Meta Model API.

## More from Meta Global AI Developer Hackathon

- [Pocket-Muse](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/680)
- [Metamorphosis](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/681)
- [Hai Tani](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/682)
- [Teesra Kaan](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/684)
- [StarAI](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/685)
- [Second Me](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/686)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/683. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt