Featured work

Projects that combine analysis, modeling, and product thinking.

This page presents selected machine learning, analytics, and application-based projects with a cleaner, more professional structure. Each project is now framed as a portfolio case card instead of a long text block.

Selected projects

I kept your original projects and links, but redesigned the layout, improved hierarchy, and converted each entry into a polished portfolio card.

Anomaly detection

Anomaly Watcher — Stock Market Anomaly Detection System

Python, Streamlit, yfinance API, stock analysis, LSTM, One-Class SVM, Isolation Forest, DBSCAN, Autoencoder

Built an advanced anomaly detection system for stock market behavior using a mix of supervised and unsupervised machine learning models.
Created a multi-page Streamlit application for real-time stock analysis, technical indicators, model comparison, and anomaly insights.
Designed a full analysis pipeline with RSI, MACD, SMA/EMA, Bollinger Bands, preprocessing, feature engineering, and evaluation metrics.

PythonStreamlitYFinanceLSTMIsolation Forest

GitHub Preview

AI productivity

Slide Genie — Text to PPT Using Gemini

Python, Gemini API, Streamlit, Google Colab, python-pptx

Developed a text-to-presentation web app that converts written content into structured PowerPoint slides with consistent formatting.
Designed an intuitive interface for content entry, slide configuration, and template selection to reduce manual presentation work.
Implemented AI-driven slide structuring and visual organization to improve presentation clarity, speed, and customization.

Gemini APIStreamlitPythonPPT Automation

GitHub Preview

Prediction system

Software Salary Price Prediction

Python, machine learning algorithms, Streamlit, Google Colab

Built a two-page salary prediction interface with separate prediction and exploration views for a cleaner user experience.
Worked on a dataset of 45K+ rows with preprocessing, feature selection, and model experimentation.
Trained and integrated models such as Linear Regression and Decision Tree into a deployable web application workflow.

RegressionStreamlitKaggle DataModel Deployment

GitHub Preview

Regression

Laptop Price Prediction

Python, Scikit-learn, regression, data visualization

Created a machine learning project to estimate laptop prices from features such as brand, RAM, storage, display, and processor.
Processed 1000+ rows of data, treated missing values and outliers, and engineered features to improve model quality.
Compared multiple regression algorithms including Linear Regression, Ridge, Lasso, KNN, Decision Tree, SVM, AdaBoost, and XGBoost.

Scikit-learnXGBoostEDAFeature Engineering

GitHub Preview

Recommendation engine

Movie Recommendation App

Python, Scikit-learn, text vectorization, similarity search, Streamlit

Built a recommendation engine using count vectorization and cosine similarity to deliver movie suggestions based on content similarity.
Integrated TMDB API for richer movie information and better presentation of recommendations.
Created a simple two-page user interface with recommendation and director-based exploration features.

NLPCosine SimilarityTMDB APIStreamlit

GitHub Preview

Computer vision

Hand Gesture Volume Controller

Python, OpenCV, Windows, VS Code, Haar Cascade

Built an early computer vision project to control system volume using hand gestures captured from the webcam.
Displayed live camera feedback, FPS, and real-time volume changes during execution.
Showcased practical experimentation with gesture interaction and OpenCV-based real-time input handling.

OpenCVComputer VisionGesture ControlPython

GitHub Preview