Diagnose chicken diseases from a phone photo.
Trained a custom image-classification model to detect common poultry diseases from a single photo, then shipped it as a mobile app farmers use in the field.
Smallholder poultry farmers couldn't access vets fast enough. By the time disease was diagnosed, a flock was often lost. They needed instant, on-device diagnosis.
We curated and annotated a labeled dataset of diseased and healthy birds, trained a CNN classifier with TensorFlow, optimized it for mobile inference, and built a clean mobile UI that gives farmers a diagnosis and recommended next steps in seconds.
Field accuracy validated against vet diagnoses. The app now works offline on low-end Android devices — exactly where it's needed.
Results
Tech stack
Got a similar challenge?
We work with founders and teams worldwide. Let's talk about your project.
Book a call