siva subramaniyen

I’m an M.Tech AI Engineering student with hands-on experience building end-to-end ML pipelines across deep learning, NLP, and computer vision. I bring 2 years of industrial engineering experience, which strengthened my problem-solving, cross-functional collaboration, and ability to deliver results independently.

I’ve built supervised deep learning and NLP systems end-to-end, including an Alzheimer’s Disease Classification pipeline using DenseNet169 transfer learning that achieved ~90% AUC. I also created a Book Recommendation System that combines VADER sentiment analysis with LDA topic modeling and is deployed as a Streamlit app.

For computer vision, I developed TextileVisionAI—a dual-model CV system using YOLOv8 for real-time tear defect detection and ResNet18 for Indian fabric pattern classification—deployed as a unified Streamlit app. I’ve also worked on Explainable AI using pixel-level attribution heatmaps to quantify each pixel’s contribution to model predictions.

In my internship, I designed and engineered prompts for an AI-powered student counselling chatbot, integrating GPT-3.5 and GPT-4 via the OpenAI API and validating performance with Postman and Python API calls. I’m now eager to apply my AI/ML expertise to real-world product innovation as a driven, adaptable team contributor.