I work at the intersection of signal processing, representation learning, and machine learning.
I am an Electronics and Telecommunication Engineering undergraduate at the University of Moratuwa, working at the intersection of machine learning, signal processing, and RF systems.
My current focus is on GNSS spoofing detection, self-supervised learning, and signal-domain representation learning, where I design systems that go from raw I/Q signals all the way to intelligent decision-making.
-
NEMESIS: Neural Embeddings for GNSS Spoofing Detection
Self-supervised representation learning framework operating on RF signals for robust GNSS spoofing detection. Accepted at IEEE ACES 2026. -
NEMESIS-Shield
Learns resilient signal representations under adversarial conditions for GNSS integrity monitoring. Under review (IEEE VTC). -
NEMESIS-Nav
Integrates learned spoofing detectors with adaptive Kalman filtering for robust navigation under attack. Under review (IEEE TVT). -
CrossJEPA
Cross-modal joint-embedding predictive architecture for efficient 3D representation learning from 2D observations. Under review (CVPR). -
SpectraNet
FFT-assisted deep learning framework for deepfake face detection leveraging frequency-domain cues. arXiv.
Business Analyst: LSEG
Worked on large-scale trade surveillance systems using Spark, Kafka, Airflow, and AWS.
Bridged product and engineering as a Product Owner.
Invited Speaker - IEEE SSCS Tech Talk
"Less is More: The Power of Small Language Models"
Email: [email protected]
Building systems where signals meet intelligence.