I build physics-informed AI systems that make spatial reasoning reliable for
real-world deployment: detecting GPS spoofing for safer autonomous navigation,
modeling energy-efficient vehicle routing, and downscaling climate data for
coastal flood-risk assessment. My research fuses deep generative models,
diffusion models, transformers, physics-informed learning, and geostatistical
priors to address data incompleteness, distribution shift, and violations of
physical laws in spatial machine learning.
I earned my Ph.D. in Computer Science from the University of Minnesota, Twin
Cities, advised by Prof. Shashi Shekhar, with committee members Prof. Vipin
Kumar, Prof. Ravi Janardan, and Prof. Ying Song. My dissertation is
Distortion-Aware Spatial Data Science (Doctoral Dissertation Fellowship,
2022-2023). At Esri, I shipped a maritime anomaly-detection pipeline on AWS that
improved classification accuracy from 55% to 73% and reduced API latency by 30%.
Actively seeking Postdoc, Machine Learning Engineer, and Research Scientist roles.
Research Statement
Teaching Statement
Diversity Statement
Ph.D. Thesis
Education & Experience
Education
University of Minnesota, Twin Cities 2018 - 2025
Ph.D. in Computer Science
Advisor: Prof. Shashi Shekhar. Committee: Prof. Vipin Kumar, Prof. Ravi
Janardan, and Prof. Ying Song. Dissertation: Distortion-Aware Spatial Data
Science. Doctoral Dissertation Fellowship, 2022-2023.
State University of New York at Buffalo 2016 - 2018
M.S. in Computer Science
Graduate training in computer science before joining the University of
Minnesota spatial computing and spatial data science research group.
Experience
Esri (Environmental Systems Research Institute) May 2023 - Dec 2023
Research Scientist Intern
-
Improved detection of route deviations and dark shipping from 55% to 73%
accuracy with an end-to-end anomaly-detection pipeline using
Transformer-based models, Evidential Deep Learning, AWS SageMaker, Lambda,
ECS, and Step Functions on roughly 500M AIS records.
-
Reduced maritime route-query latency by 40% for real-time vessel tracking
with a scalable Graph-based Traffic Representation and Association framework
built on PySpark and GeoAnalytics APIs.
-
Cut model retraining time by 35% and API latency by 30% using model
quantization, SageMaker Multi-Model Endpoints, Step Functions, SQS,
CloudWatch, and CI/CD.
University of Minnesota, Twin Cities Aug 2018 - Aug 2025
Graduate Research Assistant
-
Led Pi-DPM, a physics-informed diffusion model for detecting GPS-spoofed and
AI-generated deep-fake trajectories across maritime and urban domains.
-
Co-led Kriging-informed conditional diffusion for regional sea-level
downscaling, turning coarse climate projections into fine-grained coastal
risk maps.
-
Built GeoTrace-Agent, a multi-agent framework for auditable spatiotemporal
reasoning over AIS feeds, OSM road networks, Copernicus weather, Sentinel
imagery, and space-time-prism tools.
-
Designed Pi-GRPO, a physics-informed RL stack for trajectory generation and
trajectory-reasoning policies with PPO, DPO, GRPO, vLLM-backed rollouts, and
human-in-the-loop preference curation.