At C-DAC, I built a custom Quantum Support Vector Machine in Python and benchmarked it against classical SVMs. I also simulated Shor's, Grover's, and Deutsch-Jozsa algorithms in Qniverse.
On Q-Classify, I designed a quantum-classical spam and phishing detector that reached 94.24% accuracy with a VQC and classical SVM ensemble. The work was published in EPJ Web of Conferences at AI4SD 2026.
I built WeedReaper to detect wild radish in wheat-field images, benchmarking YOLOv11 models and reaching 85.6% test mAP@0.5 with YOLOv11n. I also co-developed Spam Shield, a BERT-based filter for SMS, calls, and email, and built a prescription-reading tool using computer vision, OCR, and NLP.

