- Leading a cross-functional team of 8 engineers, architecting scalable cloud backend and data pipeline designed to process real-time data from smart inhalers, laying the groundwork for comprehensive analysis and clinical insights.- Spearheading the development of a machine learning model for respiratory disease detection, achieving 96% accuracy on training data, with ongoing work to validate on external publicly open datasets for robust real-world performance.

- Developing a machine learning classifier for identifying respiratory diseases from audio data captured by Purcell's smart inhaler medical device, using advanced signal processing and deep learning techniques.- Architecting and implementing a cloud-based data pipeline for collecting sensor data from the smart inhaler, relaying it through the mobile app, and processing it in the cloud to provide visual analytics and machine learning-powered insights.

- Facilitated live sessions to address student inquiries related to Artificial Intelligence concepts and machine learning course projects.- Conducted comprehensive exam review sessions to assist students in revising crucial exam topics.- Proficiently resolved student queries via email and discussion forums, advocating for students' concerns with the course team.- Maintained and ensured the quality of course content and various deliverables.- Assisted the Instructor in monitoring student discussion forums, upholding academic integrity, and assisting with grading tasks.

- Graduate Services Assistant for CSE 511: Data Processing at Scale.- Guided 190+ graduate students with assignments and projects, using technologies like PostgreSQL, Apache Spark and Scala- Held regular office hours to guide students and clarify concepts of Databases and Data Processing at Scale in the course

Led the development of software architecture, backend, frontend and deployment workflows. Role included assessments, ideation, solutions, implementation as well as marketing application on Drone forums.- Developed computer vision and AI features including object detection and pose estimation using OpenCV, PyTorch, and TensorFlow. Optimized app for wide compatibility with CPU-only processing for non-GPU devices.- Created intuitive UX by integrating gamepad controls and pre-programmed flight maneuvers, reducing the learning curve and enhancing usability for users including students and old drone enthusiasts.- Integrated Google Firebase user authentication and Stripe payments for subscriptions resulting in seamless payment and subscription handling for 500+ customers.- Packaged application using Electron, simplifying product deployment across diverse user platforms.- Leveraged Flask, Nginx, and AWS EC2 to host application website, ensuring scalability and concurrent workloads.Drone AI currently has hundreds of users which include group subscriptions bought by Schools for educational purposes. The product is currently online and free to try. Visit the Product at: https://www.aidronesoftware.com

- Guided 180+ graduate students with assignments and projects, including PyTorch Collision Prediction System- Held regular office hours to guide students and clarify concepts of Artificial Intelligence in the course

- Monitored performance, query tuning and preventative maintenance on BFS client software- Trained on C#, ASP.NET and MySQL

- Designed and Developed "Change Request Portal" for accepting business requests and convert them to change requests for the concerned teams- Used Java Swing to make a standalone application with backend based on Java and SQL.