Implemented real-time video analytics systems for face detection, tracking, and anomaly detection, reduced inference latency by 40% through model optimization and quantization, and collaborated with cross-functional teams to deploy AI features into enterprise products.
Designed and deployed computer vision systems for object detection and image classification using TensorFlow and PyTorch, improving model accuracy by 15–25% through model tuning, data preprocessing, and performance optimization.
Built NLP pipelines for text classification, sentiment analysis, and chatbot applications using Transformer-based models, including BERT and GPT-based architectures.
Developed scalable data processing pipelines with Apache Spark and Python to process, transform, and analyze millions of records efficiently.
Created and maintained full-stack web applications using React, Node.js, and Django, supporting real-time analytics dashboards and data-driven user workflows.
Optimized APIs and backend services to improve system performance, reduce response time by over 30%, and support scalable application usage.