I built a Multi-Task News Intelligence System that classifies text, extracts named entities, and summarizes news across five or more categories. Using BERT-family transformers, I achieved 92% accuracy and deployed the system on AWS EC2 with an RDS database and a Streamlit interface.
For the Content Monetization Modeler, I developed a YouTube ad revenue forecasting model using eight engineered features. It achieved an R² of 0.85, and I deployed the results in an interactive Streamlit app.
I also trained a CNN to classify 12 aquatic species from more than 9,000 labeled images, achieving 94% test accuracy. I completed a B.Sc. in Computer Science with a minor in Mathematics and Statistics, and 16 data science training modules through HCL GUVI.

