Solo
Data Processor
Production-grade serverless data processing pipeline built with Python, FastAPI, and GCP services (Cloud Run, Pub/Sub, Firestore) handling 2,000+ requests per minute with 95% reliability under high concurrent load. Features event-driven architecture with fault-tolerant async processing, automatic crash recovery through Pub/Sub NACK mechanism, and exponential backoff retry strategy ensuring zero data loss. Implements multi-tenant security with Firestore sub-collection isolation, PII redaction, and comprehensive IAM controls. Fully automated deployment using Terraform Infrastructure as Code managing 15+ GCP resources and GitHub Actions CI/CD pipeline with 21 PyTest unit tests achieving 95% code coverage, format validation, and post-deployment integration tests. Achieves less than 100ms API response time with auto-scaling serverless architecture (0-10 instances), reducing infrastructure costs by 80% compared to traditional VM-based systems.
Solo
Sentiment Aura
Innovative full-stack web application that transforms spoken words into living, breathing visual art. By combining real-time speech recognition, AI-powered sentiment analysis, and dynamic weather visualization, the platform creates an immersive experience where the environment responds to the emotional tone of human speech.
Academic
End-to-End Cloud Integration & CI/CD Automation
Developed a cloud-based Health API system utilizing Python and PostgreSQL to enhance data security and efficiency for healthcare applications, while also automating cloud infrastructure provisioning through Terraform configurations for subnets, which streamlined deployment on AWS and GCP. Additionally, improved CI/CD reliability by implementing GitHub Actions pipelines with YAML, automating API testing using shell scripts, validating Packer files, and provisioning cloud instances.
Solo
Netflix Clone
Netflix Clone is a comprehensive entertainment platform built using a modern tech stack including React.js, Node.js, Express.js, MongoDB, and Tailwind CSS. It offers secure user authentication through JWT, ensuring a personalized experience. The responsive UI allows users to browse and search for movies, TV shows, and actors, watch trailers, and discover similar content. The app also features a search history function and an engaging landing page. With its combination of robust backend functionality and sleek frontend design, this application provides a seamless and feature-rich experience for movie and TV enthusiasts.
Academic
Eventify
Developed a full-stack event management platform, Eventify, featuring React Redux for efficient state management, PWA for offline functionality, and internationalization to support multiple languages, delivering an enhanced user experience. Integrated Stripe for secure payments, Google Maps for location-based event discovery, and a QR code scanner for streamlined check-ins, optimizing transaction efficiency and attendee management. Additionally, implemented real-time chat functionality using Fugu API to enable instant communication between event organizers and attendees, boosting user engagement and support.
Solo
Comic Downloader
Developed a high-performance comic retrieval application that allows users to fetch comic information via URL or text, select issues, and download images to a local directory or generate PDFs, processing over 10 comics in under a minute. The backend, built with Python and APIs and deployed using Uvicorn, seamlessly integrates with a ReactJS frontend, providing a smooth user experience for more than 100 daily active users. The application also features offline reading capabilities, storing comic names and page numbers to enable users to resume reading from their last position, enhancing overall user engagement and convenience.
Academic
Stock Market Prediction
Developed a comprehensive machine learning project focusing on sentiment analysis and predictive modeling. The project analyzed 10,000 Amazon customer reviews to determine sentiment polarity, identifying the top 50 most-reviewed products and boosting user engagement by 20%. It also conducted sentiment analysis on 5,000 IMDB movie reviews, comparing various model performances with accuracy rates ranging from 78% to 85%. Additionally, the project utilized LSTM models to predict stock market movements based on sentiment analysis of 1 million news headlines and stock data, achieving a remarkable 90% predictive accuracy that can be applied to unseen tabular data.
Solo
Todo Application
Todo Application leverages a modern technology stack to deliver a seamless and efficient task management experience. Built with TypeScript for enhanced code reliability, it utilizes MongoDB for flexible data storage, ExpressJS and NodeJS for robust backend operations, and ReactJS with ViteJS for a responsive frontend. The application's sleek interface is styled using TailWindCSS, ensuring a visually appealing and user-friendly design. With Git for version control, this project showcases a comprehensive approach to full-stack development, offering users a powerful tool to organize, track, and complete their tasks with ease.
Efficient
Innovative
Impactful
Scalable
These are my core engineering values, and I strive to embed them in all the solutions I create. I've always been passionate about building applications that simplify processes, automate tasks, and optimize performance. Solving complex problems with efficient, scalable solutions is what drives me. Collaborating with teams to deliver high-quality outcomes fuels my motivation. I enjoy architecting systems from the ground up, working across the full stack, balancing speed with precision, and continuously improving both my craft and the impact of my work. My goal is to create technology that not only empowers businesses but also makes a meaningful difference—leaving systems more optimized, users more satisfied, and processes more seamless than when I began.
I studied software engineering, starting with core programming languages like C, C++, and Java, and gradually expanding my skills into full-stack development and automation. Leveraging my strong technical foundation and problem-solving mindset, I’ve established myself as a versatile software engineer who builds scalable applications with seamless user experiences. From designing data transformation applications that went global to developing automation tools that optimized business processes, I consistently deliver high-impact solutions. Throughout my 3+ years of professional experience, I’ve worked on desktop, web, and backend applications, mastering technologies like Python, React, FastAPI, and SQL. My passion for continuous learning, combined with a collaborative approach, enables me to thrive in dynamic environments and contribute effectively to both individual projects and cross-functional teams. Now, as I pursue my master's in software engineering, I’m focused on further honing my skills in full-stack development and cloud technologies, aiming to build innovative products that solve complex problems and enhance user experiences.
Ernst & Young
Team Lead - Full-Stack Developer
Worked at Ernst & Young Global Consulting Services for 3 years, progressing from Data Analyst to Consultant, leading backend and full-stack development for data transformation, reconciliation, and document management applications using Python, FastAPI, React, Node, Angular, and SQL, optimizing performance and delivering client-specific solutions.
Zinc technologies
Python Developer Intern
Devised an RPA solution using BeautifulSoup with multithreading to crawl and extract data for 100 books in 2 minutes, storing information in a database.
The Sparks Foundation
Python Developer Intern
Completed supervised and unsupervised machine learning predictions, sports exploratory data analysis, and stock market prediction using numerical and textual analysis. Built a Dash-based UI application for IPL dataset analysis (3.7M data points), featuring 9 interactive dashboards, 12+ charts each, 20+ key statistics, search and sort functionalities, and real-time visualization customization — reducing data retrieval time by 50%.
Cyberace Infovision Private Limited
Python Developer Intern
Built a Tkinter-based UI application for a property consulting firm to enhance data visualization and decision-making for 50+ properties. Optimized SQL queries to cut data retrieval time by 36.3%, dynamically processed CSV/XLSX files to calculate rent per area for 100+ listings, and boosted user engagement by 25% with interactive visualizations for in-depth property insights.
“Mayur possesses an exceptional ability to tackle complex data manipulation tasks and can quickly make sense of even the most challenging datasets.”
MahmadSamir Khatri Tax Technology & Transformation | Ernst & Young
“Mayur's technical expertise, particularly in MERN stack development, was evident throughout the project.”
Yash Vyas Master's in Information Systems | Northeastern University
“Mayur is a star developer!”
Tapan Bansal Senior Manager | Ernst & Young
Skills
Languages: Python, Java, JavaScript, TypeScript, HTML, CSS, SCSS, SQL
Frameworks: Flask, FastAPI, Angular
Databases: MSSQL, PostgreSQL, SQLite
Tools: Power BI, Tableau
DevOps: CI/CD pipeline, Kubernetes, Docker, GitHub Actions, DigitalOcean, Terraform, Packer, AWS, GCP
Certificates
Python Programming with Project Design
Certificate of Course Completion
17 July 2020
View Certificate
Cyberace Infovision
Certificate of Internship Completion
22 July 2020
View Certificate
The Sparks Foundation
Certificate of Internship Completion
03 April 2021
View Certificate
Data Science
Certificate of Course Completion
28 March 2021
View Certificate
Deep Learning
Certificate of Course Completion
23 May 2021
View Certificate
R Programming
Certificate of Course Completion
01 June 2021
View Certificate
Tableau
Certificate of Course Completion
19 August 2021
View Certificate