Hello, I'm
Senior at Singapore American School building at the intersection of physics & artificial intelligence — founder of Sensair and Plant An Idea.
01 About
Before Singapore, I lived in Bombay. I'm really interested in physics and artificial intelligence — especially how they can be used to solve real-world problems. Most of my time outside school is spent developing Sensair, a company I founded with a friend, using physics to address clear-air turbulence.
In my free time I enjoy playing tennis and golf, exploring nature, playing the drums, listening to System of a Down, DJing, and appreciating a medium-well ribeye. Please don't hesitate to get in touch!
02 Experience
Founder of SensAir, an aviation-technology initiative addressing clear-air turbulence, one of aviation's most difficult-to-detect hazards. Led development of a real-time, onboard turbulence-detection approach integrating physics-based sensing, data-driven modeling, and machine-learning analysis — spanning applied aerospace research, system design, technical validation, and translating complex science into practical aviation solutions.
Visit sensair.net
Research assistant at the Penn GRASP Lab under PI Edward Steager, contributing to the Piccolissimo project — developing the world's smallest drone (29 mm diameter). Tested multiple frame designs and configurations, simulated flight dynamics in MATLAB, and built regression models to analyze experimental data across design iterations.
Worked with a research group to design permanent-magnet configurations imposing a Zeeman shift on barium-ion energy levels, establishing a quantization axis for optical pumping and reliable ion-state initialization. Built a magnetometer for magnet testing.
Founder of Plant An Idea, a student-led NGO using technology-driven solutions to address challenges in agriculture and education in Assam — including low-cost tools and STEM initiatives for underserved communities.
Visit plantanidea.com03 Research
Presents CatCNN, a neural network that predicts clear-air turbulence intensity by combining twelve physics-based turbulence indicators with weather reanalysis and satellite data. Trained on 150 real turbulence encounters measured by commercial aircraft, it tracked measured turbulence closely on unseen events (correlation 0.82), outperforming conventional machine-learning baselines.
Read the paper
Introduces a low-cost, modular multi-layered Cherenkov detector that identifies particles by comparing photon-yield ratios across dielectric media, eliminating the need for ring imaging. Simulations and cosmic-ray tests validate the approach, with a proposed CERN beamline test to demonstrate accurate particle discrimination across 1–15 GeV/c in an accessible, reproducible design.
04 Creations
A clear-air-turbulence sensor using doppler lidars, pulsed lasers, and machine-learning analysis.
A multi-modal AI assistant integrating natural-language processing, hand-gesture recognition, and real-time system control. Built in Python on the Orgo API to control your OS with simple commands.
A bluetooth-connected seed planter with an automatic germination sprayer, an AI scanner for bad seeds, and IR sensors for precision-farming metrics.
Built on Google Console; created algorithms to analyze data from the Piccolissimo flight testing.
05 Awards & Accolades
Whether it's physics, AI, or an idea worth planting — I'd love to hear from you.
puri46388@sas.edu.sg