o Designed and deployed a scalable, highly available, serverless AWS serverless technologies (API Gateway, Lambda, Kinesis, S3, DynamoDB, Web Sockets) to ingest over ~21,600 ML model outputs per hour from Runpod and stream actionable insights to the React frontend.

o Helped secure $1 M in pre-seed funding, enabling rapid expansion of our MVP and accelerated go-to-market for camera-free safety solutions.

o Worked at a fast-paced startup environment on Wi-Fi based user interface using React, NodeJS, Vercel enabling real-time, privacy-preserving threat detection in schools via passively interpreting Wi-Fi sensing technology—no cameras or wearables required.

o Developed and maintained infrastructure-as-code using SST (Serverless Stack Toolkit) to build new things from 0-1 quickly, accelerating cloud migrations by ~60% (using “write-once, deploy-everywhere”) and improving resource maintainability for the existing AWS infrastructure.

o Developed a 3D visualization tool using Typescript to display the Brain MRI data for Neural Connections on the web browser.

o Reading and understanding the legacy code to add additional feature to the existing platform.

o Reduced the time to fetch 1GB of data from AWS Cloud from 15 seconds to about 3 seconds, achieving an 80.91% improvement.

o Developed Boostlet.js, a JavaScript library enables real-time AI/ML processing of medical images in web browsers, created a comprehensive solution supports various data formats (e.g., EEGs, ECGs, X-rays, MRIs, and CTs).

o Designed a feature for image segmentation, data visualization, and integration with Hugging Face’s machine-learning models to enhance existing medical image processing libraries with client-side processing capabilities.

o Improved teamwork and iterative development processes using Git for Version Control, NodeJS for backend development, GitHub Actions for test automation and Submodules for external library integration.

o Implemented a clean and simple full stack recommendation platform following microservices based architecture with React as frontend framework, Spring Boot for developing a Java microservices for user management, analytics and recommendation engine.

o Implemented JSON-based user data storage with in-memory session handling and RESTful APIs for frontend-backend data exchange.

o Used transport layer like http, containerized using docker and orchestrated with docker compose for local development and integration.

o Established a testing protocol, JUnit and Jest for unit testing, and python for integration testing to validate workflows across the application.