Milind Agarwal

• Contributed to the development of a production-grade web platform for BoxCare, implementing responsive frontend
components and integrating backend REST APIs to support product catalog and business workflows.
• Collaborated with senior developers using Git-based workflows, participating in feature development, bug fixes, code
reviews, and sprint discussions.
• Technologies Used: React, Node.js, Express.js, MongoDB, JavaScript (ES6), HTML5, CSS3, Git, GitHub, Postman

• Designed and evaluated a Quantum Kernel Support Vector Machine (QkSVM) pipeline for biometric face verifica
tion using 512-D ArcFace embeddings, characterizing a peak verification accuracy of 93.74% against a near-ceiling
classical baseline of 99.18% on the Labeled Faces in the Wild (LFW) dataset.
• Architected a reproducible PennyLane simulation pipeline for 8 and 12-qubit feature maps, demonstrating a 4x scaling
improvement in True Accept Rate (TAR jumped from 0.14 to 0.57 at a strict False Accept Rate of 10−3) by scaling the
qubit and training-pair budget.
• Technologies Used: Python, PennyLane, InsightFace (ArcFace), Scikit-Learn, NumPy, Principal Component Analysis