I am designing cutting-edge algorithms and inference modules, and streamlining the workload for deep learning and labelling teams while enhancing operational efficiency.1. Engineered real-time video stream processing optimizing for efficient cloud uploads, data management reducing latency.2. Worked on computer vision applications for safety, productivity, and analysis, meeting diverse client needs.3. Crafted a custom plugin for pre-processing camera feeds, enhancing data quality for more accurate analysis4. Implemented a source switching mechanism with for ML pipeline, ensuring continuous operation and reducing downtime.1.Transform client requirements into actionable technical tasks by leading brainstorming sessions and strategy discussions.2.Conduct insightful code reviews, aimed not just at quality control but also at fostering team growth through feedback.Technical Responsibilities:1.Regularly dive into Python-based Computer Vision applications, utilizing Gstreamer and DeepStream API.2.Prioritize code optimization through consistent refactoring, leading to increased application performance.3.Author precise documentation that serves as critical guides for development and support engineers.4.Implement CI/CD pipelines to automate testing, thereby raising code reliability.5.Leverage Docker for application containerization, streamlining deployment processes.

Worked on Train ballasting detection projectalong with different objects detection, classification problems

We propose an automated real-time face attendance system where the users do not need to have physical contact with any surface. This system aims to remove the ambiguities, proxies, and ineffective nature of existing practices through a versatile solution for organizations and their members alike.This ensures a truly social distant method of attendance management in our current paradigm of virtual work. The system is powered by a Face Recognition module to authenticate the users and demonstrated up to 99.6% of test accuracy, with an improved true positive rate of 91.6% compared to existing literature.Tech Stack : Tensor flow,Deep Face, CNN, MTCNN, OpenCV

ICMS is a web application development project for the department of collegiate Education, Karnataka state government associated with Government first grade college, Raichur.Responsibilities:1. Developed and integrated the modules such as online student admission panel, Teachers attendance, Day Book, cash book, NAAC and so on.Environment: PHP, HTML5, CSS3, MySQL