I build
#Driving Innovation Beyond the Peak
CS student & AI developer building production-grade ML systems across healthcare, agriculture, fintech & road safety. 5+ years engineering intelligent solutions that matter.
I'm Job Nyaribari, a Computer Science student and AI/ML engineer passionate about building intelligent systems that solve real-world problems. From diagnosing crop diseases in remote farms to detecting driver fatigue at 95% accuracy, I engineer end-to-end AI systems that reach production.
My expertise spans the full ML lifecycle โ data collection, model training, deployment, and scaling โ with a strong focus on computer vision and predictive analytics. I'm currently deepening my Cybersecurity knowledge to build AI systems that are both intelligent and resilient to adversarial threats.
I'm open to internships, freelance collaborations, and research partnerships where I can grow and contribute meaningfully.
Every project starts with a real problem โ I build solutions that close the gap between research and impact.
I don't just train models. I deploy them โ with proper backends, APIs, and scalable infrastructure.
Focused on building AI that works for African farmers, drivers, and healthcare workers.
Currently exploring Cybersecurity to harden AI systems against vulnerabilities and adversarial attacks.
Production-grade AI systems solving real problems across multiple industries
A real-time driver monitoring system using computer vision to detect drowsiness, distraction, and unsafe driving behavior โ achieving 95% accuracy. Uses MediaPipe facial landmark detection and YOLO-based object recognition to alert drivers before accidents occur.
A full-stack AI platform empowering smallholder farmers to diagnose crop diseases via smartphone camera. Features a Django backend, EfficientNetB0 image gatekeeper, Cloudinary storage, and a micro-lending data ledger for fintech partnerships. Deployed on Railway with 500+ active users.
A specialized computer vision model for leaf-level disease classification achieving 92%+ accuracy. Trained using transfer learning on curated datasets spanning Maize, Tomato, and Potato crops with multi-class disease detection.
A diabetes risk prediction web application powered by an XGBoost model trained on clinical patient data โ achieving 78% accuracy. Features a Flask backend with a clean user-facing interface for real-time risk assessment.
An AI-driven Forex trading system currently in active development with 70%+ backtesting performance. Combines time-series forecasting, sentiment analysis, and technical indicators to generate autonomous trading signals.
Open to internships, freelance work, research collaborations, and interesting conversations.