Case Study: a software startup attracts investment with Itransition’s ML plant pathology recognition PoC

Software Startup demonstrates ML plant pathology PoC in 8 months with Itransition

The customer, a software startup, wanted to create a mobile application for plant pathology identification using machine learning. Their challenge was to develop a proof of concept that could scan, analyze, and compare plant photos to determine pathology types, a feature not fully offered by existing solutions. They partnered with Itransition for its expertise in agriculture IoT and ML-equipped solutions.

Itransition developed an ML PoC using Python and .NET, which included two neural networks for sample and pathology identification. They augmented a limited dataset and delivered a mobile app connected to a scanner and a multi-tenant web portal for experts. The PoC achieved an 80% correct pathology identification rate. As a result, the startup successfully demonstrated the solution to investors just eight months after starting and partnered with two scientific institutes for further development.