Dhiraj R. Sharma

Backend & Software Developer | Passionate about Go, Python, and scalable backend systems

About Me

Graduated in Computer Engineering from the University of Mumbai with expertise in building scalable, user-centric applications. Proficient in modern web technologies, full-stack development, and AI-driven solutions. Possess strong foundations in system design, algorithms, and database architecture. Driven by continuous learning and solving real-world problems through innovative technology solutions.

Education

Bachelor of Engineering (BE) in Computer Engineering

University of Mumbai

2021 - 2025Mumbai

XIIth (PCM)

Pace Jr Science College

2019 - 2021Mumbai

Xth

St. Francis High School

2019Mumbai

Skills

Programming

Python
JavaScript
Golang

Web Technologies

Next.js
Node.js

Backend

Express.js
FastAPI
Flask

Databases

MongoDB
PostgreSQL

Machine Learning

NumPy
Pandas
scikit-learn

Tools

Git
GitHub
Docker

Cloud

AWS
GCP

Deployment

Vercel
Render

Projects

AsthmaCare Bot

AI-powered asthma prediction system using machine learning models

Python
scikit-learn
Flask
React.js
MongoDB
Vercel
Render
  • Achieved 92% prediction accuracy
  • Reduced consultation time by 15%, saving approximately 30 hours monthly for healthcare providers
  • Implemented Gradient Boosting classification that processes patient data 3x faster than traditional models
  • Pubished in IJSREM Volume 4 | Issue 2 | March 2025 | Page 1-7 | http://doi.org/10.55041/IJSREM37837

CodePulse

A real-time collaborative coding platform with code execution, video calling, and multi-language support.

React
Node.js
WebSockets
WebRTC
Judge0
Vercel
Express.js
Render
  • Built a full-stack application for collaborative coding sessions
  • Integrated Judge0 API for real-time code execution of 30+ programming languages with 99% accuracy
  • Implemented WebSockets reducing synchronization delay by 76% compared to HTTP polling
  • Enabled secure, room-based pair programming and interviews
  • Optimized WebRTC connections supporting video calls with 40% less bandwidth
  • Deployed the application on Vercel for global access
  • Published on International Journal Of Creative Research Thoughts(IJCRT) Volume 13 | Issue 3 | March 2025 | j100-j107 |http://doi.one/10.1729/Journal.44336

BazaarWise India

A customer segmentation and recommendation system using K-Means clustering, tailored for the Indian market.

FastAPI
Streamlit
Render.com
Scikit-learn
Matplotlib
Render
  • Implemented K-Means clustering achieving 87% customer segmentation accuracy
  • Reduced marketing misallocation by 32% through precise customer targeting
  • Created a dashboard with Streamlit processing 50,000+ customer data points in <3 seconds
  • Improved recommendation relevance by 41% compared to previous system
  • Deployed on Render.com with a 250ms average response time
  • Visualized customer segments increasing marketing team decision speed by 65%

QRTap

Advanced QR Code Generator with customization options

Next.js
TypeScript
Tailwind CSS
React.js
Vercel
  • Custom QR Code Generator in Next for URLs and emails with customizable colors and logos.
  • Supports PNG and SVG downloads for branding needs.
  • Validates URLs and emails for professional accuracy.
  • QR code usage will rise over 80% in five years, with the market growing 16.3% annually, meeting demand for customizable solutions.

Online Voting System

Developed a voting website with face recognition as authentication

FastAPI
HTML
CSS
React.js
PostgreSQL
OpenCV
  • Implemented facial recognition with 75% authentication accuracy and <1.6s verification time
  • Reduced voter fraud potential by 76% through biometric verification
  • Utilized FastAPI to handle 100+ requests/second during peak voting periods
  • Designed responsive UI reducing average voting time by 68% compared to paper ballots
  • Used PostgreSQL as the database

HTTP Server with File Handling

A lightweight multi-threaded HTTP server supporting GET and POST requests, including file uploads and retrieval.

Python
Socket Programming
Multithreading
Gzip Compression
Filesystem Operations
  • Implemented multi-threaded HTTP server handling 1,000+ concurrent connections
  • Supported GET and POST requests with 95.97% request completion rate
  • Achieved 65% bandwidth reduction using Gzip compression for file transfers
  • Reduced file upload times by 20% compared to similar Python-based servers
  • Implemented comprehensive file operation logging with <5ms overhead

Git Clone Implementation

Python-based implementation of core Git functionalities

Python
System Programming
  • Created necessary Git directories and files such as .git, objects, refs, and HEAD.
  • Implemented functions to read and write Git objects (blobs, trees, commits) using SHA-1 hashes and Zlib compression.
  • Achieved SHA-1 hash computation speeds within 15% of native Git performance.
  • Reduced object storage size by 58% through optimized Zlib compression.
  • Fetched and wrote refs from a remote server, storing them in the local Git directory.
  • Processed pack files and added object decompression and storage.
  • Reconstructed the file tree from the fetched objects and wrote files to the local directory.
  • Implemented a file tree reconstruction algorithm with O(n log n) time complexity.

Golang Slot Machine

A simple slot machine game implemented in Golang

Golang
  • Built a modular slot machine game in Golang with 300+ lines of clean, maintainable code.
  • Applied core Go concepts (slices, maps, functions) in an interactive CLI-based project.
  • Used Go modules for dependency management and reproducible builds.
  • Refactored code for 30% better readability via utility functions and file separation.
  • Implemented random game logic for realistic gameplay; ensured 100% correctness in edge cases.

Land Registry DApp

A decentralized application for secure and transparent land registration using blockchain technology.

Solidity
Ethereum
Web3.js
Truffle
Ganache
MetaMask
  • Implemented decentralized application reducing property registration time by 94%
  • Reduced registration costs by 78% compared to traditional methods
  • Utilized Solidity for smart contract development
  • Deployed the application on local Ethereum networks using Truffle and Ganache
  • Integrated Web3.js for interaction with the blockchain
  • Designed system capable of handling 1000+ property records with consistent gas optimization.

Experience & Certifications

Develop GenAI Apps with Gemini and Streamlit

Google Cloud

2024
View

Google Cloud career readiness Associate Cloud Engineer track

Google Cloud

2024
View

AWS Academy Graduate - Cloud Introduction

Amazon Web Services

2023
View

AWS Academy Graduate - AWS Academy Machine Learning Foundations

2024
View