Harsha Kota

I’m a B.Tech AI & ML student with hands-on experience in computer vision and end-to-end machine learning solution development through internships at Infosys Springboard and AICTE Microsoft–SAP.

I’ve built real-time object detection systems using YOLO, Faster R-CNN, and Mask R-CNN, and I focus on turning models into deployable products.

In my Infosys Springboard internship, I developed an end-to-end Object Recognition System using Faster R-CNN, Mask R-CNN, and YOLOv5, achieving 95% accuracy on ~5,000 images. I also created a Streamlit-based real-time detection application for image and video processing, improving inference time by 20%.

I care about robust pipelines and practical deployment workflows—data preprocessing, model evaluation metrics, and production-ready deployment using Streamlit and Flask. I also built WildDiscover, a React Native app for real-time wildlife and landmark identification using multiple APIs and AWS S3 for secure image storage.