HELLO! I AM
Data Science Undergraduate | AI & ML Enthusiast
I'm passionate about developing efficient, scalable solutions to streamline complex tasks and address data science challenges.
Take a look at some of the things I love working on.
Turning raw data into clear, actionable insights.
Cleaning and transforming raw data into useful model features.
Making sense of text with sentiment, topics, and classifiers.
Building and tuning models that actually work.
Creating smart “you might like this” suggestions.
Training neural nets and fine-tuning transformers.
An AI-driven platform using LLMs and NLP to personalize learning content from YouTube, Coursera, Udemy, Wikipedia, and arXiv. Features semantic search via Pinecone, secure JWT authentication, and real-time WebSocket delivery for tailored educational experiences across all learner levels.
Analyzes emergency call audio for stress, emotions, speaker diarization, and critical incidents using SER and APIs. Scores dispatcher performance, assesses threat levels, extracts key info, and generates interactive visualizations and reports to enhance dispatch center efficiency.
Processes PDFs, text, and HTML to extract entities/relations via BERT NER, builds Neo4j graphs, enables reasoning with NetworkX, and offers interactive Streamlit visualizations for exploring and querying extracted knowledge structures.
ML-based model predicts telecom churn with feature insights via an interactive Streamlit app, enabling data analysis, visualization, and proactive retention strategies for customer datasets.
Comprehensive ERP web app for Wijesundara Rice Mill, handling inventory, orders, transportation, accounting, HR, CRM, and suppliers. Includes custom tools for stock tracking, task assignment, bidding, and real-time operational efficiency across departments.