At Nass Technologies, I worked on a Student Administration & Online Test Platform, analyzing student and examination data to generate reports and performance insights.
I designed and optimized SQL Server databases, cleaned and validated data, and created dashboards to track student performance and test analytics. I also worked with the development team to analyze business requirements and improve data accuracy and reporting efficiency.
On my video sentiment and emotion analysis project, I developed a deep learning pipeline that achieved 92% accuracy across multiple classes. I optimized training with PyTorch and AWS SageMaker, reducing training time by 30%, and built an interactive Gradio interface for predictions and visualization.
I also built and deployed a hand sign recognition CNN for 25 gesture classes, achieving 95% accuracy. The project used image preprocessing and OpenCV pipelines for real-time classification.
