Mayank Prakash Sharma

CALIFORNIA STATE UNIVERSITY – DOMINGUEZ HILLS

GPA: 3.5/4.0

NAVRACHANA UNIVERSITY

• Architect and maintain a distributed microservices-based shipment-tracking system using Node.js, Docker, and AWS Lambda, processing 10,000+ shipments per month, and implement observability for rapid diagnostics.
• Develop and enhance the Next.js SSR application integrating Firebase Firestore and Cloud Functions to secure API credentials and improve page-load performance by 40%.
• Integrate and synchronize data from Amazon Seller Partner, Etsy, Walmart, FedEx, UPS, and ShipStation APIs into a unified order-management dashboard, handling 5,000+ weekly orders with 95% accuracy.
• Design and enforce Firestore Security Rules and Role-Based Access Control (RBAC) to govern data access for 20+ admin users, maintaining 100% compliance with internal security policies and security best practices.
• Optimize Firestore database performance through composite indexes, denormalization, caching strategies, and performance testing, cutting query latency by 64% and saving $5K monthly in cloud costs.
• Implement and maintain CI/CD pipelines with Jenkins and GitHub Actions for automated testing and deployment, applying DevOps practices to reduce release cycle time from 3 hours to 30 minutes and achieving zero downtime.
• Monitor production systems, troubleshoot issues, and roll out hotfixes to ensure 99.9% application uptime, reliability, and rapid incident resolution.

• Engineered a full-stack mobile application in Flutter and Kotlin, managing the entire SDLC (design, development, testing, and deployment), reducing time-to-market by 20%.
• Collaborated with cross-functional teams, gathering requirements from 50+ stakeholders, and designed user interfaces to enhance customer experience by 30%.
• Developed and integrated RESTful APIs and third-party libraries into the Node.js backend, improving system performance and enabling real-time data handling with Firebase while ensuring scalability within virtualized environments and distributed systems.

• Developed and maintained consumer-facing websites and mobile applications for 13k+ users, leveraging operating systems fundamentals and ensuring scalability and high reliability.
• Designed and implemented a web scraping tool in Java and Python to automate data collection for 250+ school clubs, increasing data accuracy by 20%.
• Led cross-functional collaboration efforts to integrate third-party services and ensure system scalability, reducing downtime by 15% through proactive maintenance.

Real-Time Automatic Number Plate Recognition (ANPR) System

• Built and deployed a Python-based web scraping + preprocessing pipeline to collect and clean 10,000+ vehicle images for training.
• Trained an image detection model using YOLOv2 in TensorFlow, improving robustness through extensive training and hyperparameter tuning.
• Designed a custom OCR pipeline using adaptive thresholding, binarization, and smoothing to extract license plate characters accurately.
• Optimized real-time inference performance and integrated the solution into a web application for scalable deployment.

AI-Driven Vibration Analysis & Predictive Maintenance Platform

• Developed an AI-powered platform for anomaly detection, remaining useful life (RUL) prediction, and root cause analysis for mechanical equipment (pumps, compressors, fire engines).
• Preprocessed and labeled time-series vibration data using filtering, curve fitting, and statistical learning methods with research and data teams.
• Implemented and benchmarked 21+ models (regression, LSTMs, autoencoders), selecting and tuning top performers for production.
• Built an in-house AutoML retraining workflow to reduce model drift and support deployment with IoT sensors and real-time dashboards using MLflow and PyCaret.

• Developed a full-stack loan management system with Node.js and Firebase, handling real-time financial data while ensuring scalability with Google Cloud.
• Built RESTful APIs for notifications and automated late fee management via Google Cloud Functions and Nodemailer.
• Collaborated with cross-functional teams to optimize the payment workflow, improving efficiency by 20% through backend enhancements in Node.js and Firestore.

View Project

• Uncovered new genetic variants and analyzed public health implications using a dataset of 9,948 full SARS-CoV-2 sequences.
• Created an LSTM neural network model to study links between viral alterations and epidemiological outcomes, analyzing 1,000+ samples with Cedars-Sinai.

• Designed and developed a rewards management web app using MERN (MongoDB, Express.js, React.js, Node.js) with Firebase authentication for secure user management.
• Implemented CI/CD pipelines and deployed using AWS Elastic Beanstalk and Amplify for scalable, reliable performance.
• Built RESTful APIs to manage rewards data, integrating third-party services and ensuring a seamless user experience.

View Project

• Engineered a smart mirror displaying time, date, weather, news, RSS feeds, maps, and other details using a two-way mirror panel, display monitor, Raspberry Pi, and desktop application.
• Implemented Raspberry Pi and Node.js features by extending open-source modules and delivering a usable end-to-end prototype.