At Prinston Smart Engineer, I directed the end-to-end ML pipeline, from data preparation and model training through evaluation and deployment in interactive Streamlit interfaces. I trained supervised learning models and built ANN and CNN architectures with TensorFlow and Keras.
I built a Flask application that uses a custom CNN to detect mulberry leaf and silkworm diseases from uploaded images, with diagnosis and treatment recommendations in English and Kannada. I also deployed an earthquake magnitude prediction model as a Streamlit app and built a Python and Django student management platform.

