Web Projects

No-Code Web Builder

A drag and drop website builder using the MERN stack and Tailwind CSS. This tool allows users to effortlessly build responsive websites by dragging and dropping components. Once designed, users can obtain the corresponding code to implement their creations.

Exam Preparation

Used Next.js 14, Prisma, Postgres, Tailwind, NextAuth v5, TypeScript and ShadCN. This platform enables users to prepare for the JEE examination by solving practice questions and participating in proctored exams along with learning roadmaps. It boasts impressive scores on Google Lighthouse with 100 in performance and SEO.

Algo Visualizer

• React-based algorithm visualizer showcasing various algorithms like searching, sorting, path-finding, etc. • Implemented various algorithms, such as selection sort, binary search, and BFS, using React. • Includes AI Algorithms like A* which takes 6X less time than Dijkstra.

AI Projects

End-to-End ML Pipeline

This end-to-end machine learning project predicts student performance, incorporating components such as data ingestion, transformation, and model training. It features train-test pipelines, logging, exception handling, and a user-friendly Flask interface which lead to a R2 score of 0.8809

Data Preprocessing Interface

This project provides an interface for data preprocessing and visualization, including missing value handling, scaling, encoding categorical features, outlier management, feature selection, and diverse graph plotting. It ensures efficient data preparation and insightful visualization for enhanced data analysis.

Makemore (Andrej Karpathy)

Inspired by Andrej Karpathy's Neural Network Zero to Hero, this project consist making of a character level language model. It begins with a bigram character-level language model. The series covers torch.Tensor, MLP models, deep network internals, Batch Normalization, and progresses to a WaveNet-like structure.

ML in Finance

CAPM

This project implements the Capital Asset Pricing Model (CAPM) using Python and various libraries such as pandas, statsmodels, and matplotlib. The CAPM is a widely used model in finance that describes the relationship between the expected return of an asset and its systematic risk.

Pair Trading

This project implements a pair trading strategy using historical stock data. Pair trading is a market-neutral strategy that involves taking long and short positions in two highly correlated stocks. The strategy bets on the mean-reverting behavior of the price spread between the two stocks.