Technical projects spanning machine learning, systems design, and developer tools. Each focused on solving real problems with clean, scalable solutions.
Research · ICAR-IARI, DRDO, IIRR
I developed a Convolutional Neural Network to detect Rice Blast disease in paddy crop before onset of physical symptoms. The model was trained on cleaned experimental data collected from ultrasonic emissions of healthy and diseased specimens. The model was developed in TensorFlow and achieved an accuracy of 99.12 percent on test data.
Research · Sustainability
I developed low-cost, modular water filters using plant xylem, granular activated carbon, and hemp fibers. The design focused on sustainability and accessibility for rural communities. I presented my work alongside my teammate at the Stockholm Junior Water Prize at IIT Madras. A prototype was implemented in Bambawar Village, positively impacting over 1000 residents by providing cleaner drinking water.
IoT · Environmental Tech
I developed a kit that uses sensors to measure 20+ water quality metrics, both physical and chemical. Using IoT technology, I connected the kit to mobile app to display real time data. Collaborated with NITI AAYOG to deploy the system in Abupur and Bambawar villages for pond water monitoring, and taught 100+ pond ambassadors how to use it.
Social Justice · Tech
Developed a platform that scrapes government schemes from multiple sources and uses multiparamater filtering to match users with relevant schemes to their needs. The platform has impacted over 3.5k users and reached more than 20k users.
Social Impact · Antigone SIFP
Used the MapBox API to build an interactive pond mapping and data collection platform. The platform allows users to upload water quality data and images to create a unified pond database. Secured a ₹15k grant for development, and collaborated with NITI AAYOG for data collection.
Work · Dimension
Led team of 3 developers. Developed a reactive dashboard, responsive to real-time data and all view ports. Integrated various third-party APIs to enhance functionality. Worked at US-based startup ($2M+ funded).
Hackathon · Sustainability
Carbon footprint tracking website helping users monitor and reduce their environmental impact by gamifying sustainability. The app used Next.JS and Express JS to create a full stack real-time experience. 500$ Cash Prize and Runner-up for Best Beginner Project at HackTech 2023.