Hello! I'm Sharath Tunuguntla.
Applied ML Engineer with 6+ years of end-to-end experience building, training, and deploying computer vision and AI systems in real-world production environments. Specialises in object detection, Re-Identification (ReID), tracking, and GPU-accelerated inference pipelines — with equal depth in the MLOps infrastructure required to run them reliably at scale across hundreds of remote sites. Comfortable from model training through to Docker-based deployment, automated rollback, and live GPU resource management.
My Expertise
I can provide real-time data analytic solutions for many vision based AI usecases. And I am pretty skilled at web-design as well.
Click on the following icons to know more about my expertise:
My Skills
I am a quick learner and specialize in multitude of skills required for Model Training, Pipeline Building, Web App Development and Product Design
Experience
Principal Software Engineer
Built real-time computer vision analytics across 80 camera streams. Developed ReID with SigLIP, integrated vLLM to cut QA effort from 50 to 15 hrs, and built an automated deployment service managing 400+ remote servers.
Created a QA tool for the ReID process that allows QA team to browse through the images and check reID results.
Machine Learning Developer
Developed and maintained production AI pipelines for real-time computer vision — including traffic counting, pose estimation, employee interaction tracking, and checkout analytics — using YOLOv4, YOLOX, ResNet, and DeepStream.
Designed and Migrated platform to microservices.
Software Engineer
Built backend web applications using Python/Django on AWS. Developed a loan filtering algorithm, set up REST APIs, managed Docker/Celery/Jenkins deployments, and refactored large codebases for maintainability.
Worked in a team with sprint based development and agile methodologies with 20+ developers.
Miscellaneous Projects
Built CV systems for baggage detection, parking alerts, traffic heat maps, and employee interaction tracking using YOLO and DeepStream.
Also developed a Deep Q-learning navigation agent, a speech recognition system using RNN/CNN, a robotic arm kinematic simulator, and a personal productivity web app.
Testimonials
Contact Me
If you like my work and want to hire me then drop me a message using the contact form.
Or get in touch using my email or my contact number.
See you!