At Fxis.AI, I built a Python pipeline that extracted structured candidate information from resumes using the ChatGPT API, achieving approximately 85% accuracy in manual evaluation. I also designed prompts and validation logic to handle different resume layouts without filling missing information with guesses.
At Fxis.AI, I developed a recommendation system using SQL queries and ranking logic that increased recommendation engagement by approximately 20–25%. I also integrated the Twilio WhatsApp API with the recruitment workflow to automate candidate communication.
At Prodigy InfoTech, I developed a PyTorch convolutional neural network for hand gesture recognition. I contributed to a vision system that classified gestures with over 90% test accuracy and helped integrate and validate the model in the application.
My projects include a deep learning pipeline for gait recognition and classification using the CASIA-B dataset, with over 85% test accuracy. I also developed a water-potability classification pipeline in which XGBoost achieved 93.07% accuracy; I’m pursuing a Master’s in Computer Science at Arizona State University.
