At MNR Technologies, I built ClipCut, an AI video-editing pipeline for YouTube-optimized thumbnail generation and natural-language-prompted edits, including auto zoom-ins and silence or blur removal.
I also built an LLM-driven QA framework using Playwright, GPT-4o, and Claude. It profiles DOM and vision elements, repairs XPath selectors, and generates QA summaries and Excel reports.
At Unified Mentor Pvt. Ltd., I developed and deployed end-to-end machine learning applications in Python for NLP, regression, and classification use cases, working with senior mentors. Earlier, I implemented NLP and regression and classification models for case studies at Bharat Intern.
My projects include AgriAI, a crop-health diagnostic tool using TensorFlow and YOLOv5, and a phishing URL detector deployed with Flask. I also built MindTrace AI, a stress-detection agent that reached 85% emotion and micro-expression accuracy.
