I'm Building Evotics. In the past, I’ve built a script-to-music brief system, worked on a multilingual dubbing pipeline, and implemented a DistilBERT-based classifier.

Previously:

• Designed and trained CNN-based models for stroke risk prediction and biomedical image classification.
• Built E.C.H.O, an end-to-end AI pipeline for multilingual video dubbing and subtitling using speech recognition and voice synthesis.
• Fine-tuned DistilBERT on 37,000 IMDb reviews for spoiler detection, achieving strong NLP performance.
• Created Fible, an NLP tool converting movie scripts into music production briefs using emotional analysis.
• Research intern at IIIT Kottayam worked on LLMs, NLP, and computer vision applications.

Publications:

• Enhanced Drunkenness Detection via Iris Pattern Classification with Optimized Pre-Trained Deep Learning Models — IEEE 4th International Conference on Advances in Computing, Communication, Embedded and Secure Systems (ACCESS'25), 2025.