Experience

• Built multi-sensor volumetric capture system with 4 Azure Kinect cameras and RabbitMQ messaging, achieving real-time 360° reconstruction at 30 FPS.
• Developing federated learning framework (MetaFL) for privacy-preserving biometric authentication, improving accuracy from 93.8% to 94.3% through hard negative sampling and pre-trained models while scaling to 50,000+ user datasets.

• Architected file management application using Svelte, Tauri, Python, and Rust, processing 10,000+ files and reducing search time by 45 minutes weekly per user.
• Implemented ML classification pipeline achieving 80% accuracy on file categorization with sub-second processing for 500+ file batches.

• Tutored 25+ CS students with personalized instruction, achieving 99% satisfaction rating and 18% average grade improvement.
• Created visual learning materials and interactive exercises for data structures and algorithms, with 80% of students reporting improved problem-solving abilities.

• Supported 600+ students in C, Java, and Python through office hours and code reviews, improving assignment completion rates by 15%.
• Led 20+ review sessions on data structures and algorithms, with participants demonstrating improved performance on examinations.