All Projects
Every project I've built — machine learning pipelines, data dashboards, and full-stack apps.
CoolRoute
AI-powered route planner that finds the walking, riding, or driving route with the least heat exposure not just the shortest one. Built for the FortyGuard Global AI Hackathon 2026 using hyperlocal temperature intelligence + real routing data.
AI Fashion Recommendation System
Upload a clothing photo and get visually similar items back instantly. Uses a pretrained PyTorch ResNet50 as a feature extractor (transfer learning, no training needed), FAISS for millisecond similarity search across a 2000+ item catalog, and a full Streamlit UI with gallery, dashboard, favorites, and history.
QuickBite — AI Food Ordering Chatbot
An NLP-powered food ordering system built with Dialogflow for intent recognition and entity extraction, backed by a Python/Flask service and MySQL database for order tracking, menu management, and live order status.
NLP Hub
A single web app bundling six focused NLP tools summarization, sentiment analysis, text generation, extractive Q&A, named entity recognition, and text similarity each powered by its own pretrained Hugging Face model, served through a FastAPI backend with a Next.js frontend.
NLP — Learning Series
A structured, hands-on journey through NLP: regex text cleaning, tokenization, POS tagging, NER, stemming/lemmatization, bag-of-words, TF-IDF, word embeddings, Word2Vec, LDA topic modeling, and FastText all in runnable, self-contained notebooks.
LangChain & GenAI — Learning Series
Hands-on learning repository covering the LangChain ecosystem end-to-end: LLMs, chat models, prompts, structured output, embedding models, vector stores, retrievers, document loaders, runnables, chains, tools, and agents.
Deep Learning with TensorFlow & Keras
A hands-on collection of deep learning notebooks from how a neuron actually learns (forward/backprop, gradient descent) up through CNNs, RNNs, LSTMs, and Transformers. Includes MNIST digit classification, Dogs vs Cats transfer learning, LSTM sequence forecasting, and a from-scratch Transformer implementation.
Machine Learning — Learning Series
Ground-up, hands-on machine learning notebooks: feature engineering (ordinal & one-hot encoding, feature scaling standardization and normalization), regression, classification, and ensemble methods, each in its own self-contained folder.
Machine Learning Projects (5-Project Collection)
A growing collection of beginner-to-intermediate ML projects, each with its own notebook, dataset, and documentation: Sonar Rock vs Mine classification, Diabetes Prediction, House Price Prediction, Fake News Detection, and Credit Card Fraud Detection.
Credit Card Fraud Detector
ML dashboard detecting fraud using 6 supervised learning algorithms with interactive charts and real-time predictions.
Movie Recommender System
Streamlit-based recommendation engine using cosine similarity and TMDB API.
SpaceX Launch Success Prediction
Predicted Falcon 9 landing success using Logistic Regression, SVM, Decision Trees. Interactive Plotly Dash dashboard.
Data Visualization Dashboards
Australian wildfire activity and automobile sales trends dashboards.
SQL with Python — Learning Series
Hands-on practice notebooks for databases & SQL with Python: SQLite fundamentals, SQL Magic in Jupyter, querying real-world datasets, and a full end-to-end project analyzing Chicago public datasets with SQL.
Python Basics — Learning Series
A beginner-to-intermediate collection of Jupyter notebooks covering core Python: variables, input, strings, math operations, conditionals, lists, tuples, dictionaries, functions, exceptions, and classes one focused notebook per topic.
ZenWallet
Expense tracker with budgets, goals, charts, currency conversion.