For the disease classifier, I programmed a CNN for plant disease classification along with fine tuning a Gemini LLM to output advice on plant health with RAG. This project won the UC Davis Hacks 2024 Intel prize, probably the most competitive prize. For the harvest time recommender, I worked with a plant science PhD student who had some very useful knowledge about the needs of the industry. In this particular case they needed to know the size of spinach plants across a field. I used Google Cloud to run pre-trained image segmentation model, and trained a transformer model to classify images.