Hi, I’m Hariharan

I’m an aspiring AI/ML engineer with a strong interest in NLP, deep learning, and Explainable AI. I enjoy creating intelligent models that are both accurate and interpretable. My recent MSc project focused on Aspect-Based Sentiment Analysis using BERT and LIME for explainability. I’m actively seeking full-time opportunities to apply my skills in solving real-world AI challenges.

Technical & Professional Skills

⭐ Natural Language Processing (NLP) | 3 yrs

⭐ PyTorch | 4 yrs

⭐ Streamlit | 1 yrs

⭐ Transformers / BERT | 1 yrs

⭐ Python | 4yrs

🤖 Featured Projects

Aspect-Based Sentiment Analysis using BERT and LIME

Python | PyTorch | BERT| LIME | Streamlit

Description: Developed an ABSA model to extract aspect-level sentiments from product reviews using fine-tuned BERT. Implemented LIME for interpretability, enabling users to visualize why specific sentiments were predicted. Deployed as a Streamlit app for real-time input and result visualization.

Aspect-Based Sentiment Analysis using BERT and LIME

Smart Study Organizer – AI-Powered Desk Detection System

Python | YOLOv5 | OpenCV | Flask | HTML/CSS

Description: Developed a real-time object detection tool that scans and identifies study items (notebooks, pens, laptops, etc.) using a webcam. Built with YOLOv5 and OpenCV, and deployed via a Flask web interface, the system encourages workspace organization and boosts productivity.

Smart Study Organizer

Tiny Object and Edge Detection Model

Python | YOLOv4 | OpenCV | TensorFlow | Canny Edge Detection

Description: Engineered a deep learning pipeline to detect tiny objects and fine edges in cluttered images. Utilized YOLOv4 for small-scale object recognition and integrated OpenCV's Canny edge detection to highlight structural image features. Ideal for precision tasks in low-visibility or complex backgrounds.

Tiny-Object-and-Edge-Detection-Model

Screenshot (35).png