Ijisho: offline bean disease check
Spots angular leaf spot and bean rust from a phone photo, offline, in English and Kinyarwanda. 92% accurate on unseen field photos. Open source.
Try the app
Research and open-source work beyond client websites, mostly for agriculture, education and public services in Rwanda.
The code for each project is public on GitHub.
Spots angular leaf spot and bean rust from a phone photo, offline, in English and Kinyarwanda. 92% accurate on unseen field photos. Open source.
Try the appA Raspberry Pi device that recommends crops from soil nutrients and weather data for Rwandan farmers.
View on GitHubA marketplace connecting farmers and customers, with location-based discovery and price comparison.
View on GitHubVehicle detection from several cameras with YOLO, driving NodeMCU-controlled traffic lights.
View on GitHubAn Arduino gas sensor with a Flask dashboard that logs readings and raises alerts.
View on GitHubA practice platform built on NESA past exams, with NLP answer grading and PDF question extraction.
View on GitHubCompiled landslide datasets (CSV, GeoJSON, KML), maps and risk-assessment reports for Rwanda.
View on GitHubRoofing materials store with AI recommendations, sentiment analysis and a roof cost calculator.
View on GitHubA farmer marketplace with Solidity smart contracts, built and tested with Hardhat.
View on GitHubMicroservices with Docker, Kubernetes, Jenkins and a GitHub Actions CI/CD pipeline.
View on GitHub
Tell me what you want to build. I’ll reply with questions and a plan for a first version.