Real-world projects I've built
Built a production-ready hybrid recommendation system combining ALS Matrix Factorization (collaborative filtering) with TF-IDF + Cosine Similarity (content-based filtering). Trained on MovieLens dataset (3,238 users, 21,509 movies, 0.72% matrix density). Features an interactive Streamlit dashboard with real-time personalized recommendations and tunable hybrid weight (α).
🔗 View on GitHub →Built an AI system capable of processing multiple input types (text, image, etc.) for versatile real-world use. Integrated research assistance and data analysis features, enabling intelligent insights across different data formats. Implemented cross-modal understanding to support complex queries and automated analytical tasks.
Developed a regression model to predict house prices based on key features like location and size. Applied data preprocessing, feature engineering, and model evaluation techniques to achieve accurate predictions.
Created a voice/command-based desktop assistant to automate daily tasks. Integrated features like web search, file management, and system controls for enhanced productivity.