Srujan B

I’m an enthusiastic, detail-oriented computer science student at KLE Technological University with a strong foundation in machine learning, deep learning, and computer vision, paired with hands-on software engineering experience. I’m motivated by using technical depth to solve real problems and deliver engineering-quality outcomes.

In my Computer Vision internship at Arctictern Solutions, I contributed to AnnoFlow, a full-stack CV annotation and model training platform, and worked directly with production-scale dataset preparation for autonomous vehicle perception pipelines. I helped build annotation and review workflows, performed QA validation to improve label accuracy and consistency, and supported model training at scale using async pipelines and artifact versioning.

I also reinforce my skills through projects that blend predictive modeling and deployment thinking, from detecting insert rejections and building RFID-based asset tracking to training CNNs for defect detection in cultural heritage sites using interpretable Grad-CAM localization. With published conference work in deep learning for heritage buildings and BERT-based sentiment analysis, I’m eager to keep pushing from experimentation to dependable, real-world systems.