real-time product recommendation engine using collaborative filtering (ALS) with online learning. combines batch-trained embeddings with real-time user profile updates via exponential moving average for personalized recommendations.

an automated machine learning pipeline that takes you from raw data to a production-ready model with just one command.

full-stack car price prediction web app. complete ML pipeline from data scraping to model deployment. django rest api backend, next.js frontend, catboost model, docker containerization.

langchain-powered multi-agent ai debate app with fastapi backend and next.js frontend. supports openai, google gemini, and ollama for local models with real-time streaming via websockets.

a multi-agent system project built with Google ADK that conducts personalized technical mock interviews by analyzing your resume and target job description.

movie recommendation system implementing popularity-based, content-based, collaborative filtering, and hybrid approaches.

fashion MNIST image classifier using tensorflow/keras. includes complete workflow and pre-trained model.

a pytorch implementation of deep convolutional generative adversarial network (DCGAN) for generating images.

evolutionary algorithm that recreates images using overlapping geometric shapes.