Birendra Sharma

Sticky Notes

A web application developed using Django, HTML, and CSS that allows users to post blogs and share stories related to Machine Learning and Data Science topics.

Live Demo

• Built with Django framework for backend functionality
• Created responsive UI using HTML and CSS
• Implemented user authentication and authorization
• Designed database schema for storing and retrieving blog posts

Multiplayer Bingo (Rooms + WebSockets)

Real-time, room-based multiplayer Bingo built with React + Vite (frontend) and FastAPI WebSockets (backend). Features turn enforcement, deterministic boards per player, and multi-winner draw handling.

Live Demo

• Create or join rooms using 5-character codes
• Real-time sync via WebSockets with automatic reconnection
• Turn-based gameplay with server-side enforcement
• Deterministic boards (seeded by room ID and player name)
• Draw support when multiple players win simultaneously
• Modern, responsive UI with accessible focus and animations
• Deployment: Frontend on Vercel, Backend on Render

ChatQwen

A web application built with Streamlit, serving as a personal ChatGPT for answering basic coding and general queries. It leverages an LLM model fine-tuned on a Python queries dataset to provide accurate and context-specific responses.

• Developed using Streamlit for quick and effective UI development
• Integrated fine-tuned LLM model for answering coding queries
• Implemented context-based response generation
• Optimized for handling Python-specific programming questions

Heart Disease Prediction

A project developed using Tkinter, scikit-learn, pandas, and Matplotlib to predict whether a patient has heart disease based on their current medical details.

• Built GUI interface with Tkinter for easy data input
• Implemented machine learning models using scikit-learn
• Performed data analysis with pandas for preprocessing
• Created visualizations using Matplotlib to present results

MCQs Generation using NLP

An NLP project developed using Django, Transformers, and a fine-tuned LLM. This web application allows users to input text and generate multiple-choice questions (MCQs) along with distractors based on the provided content.

• Built with Django framework for robust backend functionality
• Integrated Transformer models for NLP processing
• Implemented fine-tuned LLM for question generation
• Created algorithms for generating relevant distractors

Crop and Fertilizer Recommendation System

An ML-powered web application developed using Flask, HTML, CSS, scikit-learn, pandas, and MySQL. Users can input specific data, and the system provides recommendations for suitable crops and fertilizers based on the input parameters.

• Developed full-stack application with Flask backend and HTML/CSS frontend
• Implemented machine learning algorithms using scikit-learn for accurate recommendations
• Created a MySQL database for storing crop and fertilizer data
• Designed user-friendly interface for data input and result visualization