---
title: "Swiper No Mauling"
startDate: "2019-02-01"
endDate: "2019-02-01"
description: "Engineered a financially viable, AI-powered computer vision application designed to help farmers autonomously monitor livestock and prevent predator attacks."
repo: "https://github.com/prdlk/CV-Livestock-Protection"
tags: ["Python", "TensorFlow", "YOLOv3", "Computer Vision", "Machine Learning", "Twilio API"]
---

- Engineered a financially viable, AI-powered computer vision application designed to help farmers autonomously monitor livestock and prevent predator attacks.
- Implemented the YOLOv3 object detection model using Python and TensorFlow to accurately identify predators and assess hostility levels within live camera frames.
- Integrated the Twilio API to build an automated, real-time alert system that immediately dispatches SMS notifications to farmers the moment a threat is detected.
- Designed an architectural roadmap to transition the computationally heavy detection model to a cloud-based infrastructure, with planned expansions for wildlife researchers to track events of interest.