Lafayette, LA · Open to AI/ML & Geospatial roles

Turning pixels, satellite, UAV, and medical images into decisions.

I'm an AI/ML engineer working across computer vision, remote sensing, and language models. I've trained segmentation and detection models on more than a dozen architectures, shipped geospatial ML pipelines used in production, and published research on transformer-based wildfire detection.

LOG>
Selected work
Four projects spanning medical imaging, disaster response, fire detection, and municipal GIS.
Computer VisionResearch

Fine-tuning SAM for abdominal organ segmentation

Adapted Meta's Segment Anything Model for prompt-based, multi-class segmentation of abdominal MRI organs — freezing the encoder and training a custom decoder for the CHAOS dataset.

PYTHON · PYTORCH · SAM · MONAI · WANDB
Computer VisionGeospatial

Post-disaster rapid damage assessment from UAV imagery

Built an assessment pipeline for disaster recovery teams — trained YOLO on labeled UAV datasets and generated damage hotspot maps for on-the-ground planning.

YOLO · GEONADIR · CVAT · ROBOFLOW · S3
Computer VisionGeospatialResearch

Wildfire detection from Landsat-8 imagery using ViT

Framework for fire segmentation across satellite imagery, benchmarking ViT-TransUNet, MedT, and U-Net. Presented at CVCI 2025 and published in ACM Proceedings.

LANDSAT-8 · VIT · TRANSUNET · U-NET · ONECYCLELR
GeospatialLLM & NLP

Text-to-SQL for a municipal GIS web application

Built and deployed a production ML pipeline that lets Tokha Municipality staff query GIS databases in plain language, translated to SQL via LangChain and LangGraph.

LANGCHAIN · LANGGRAPH · POSTGRESQL · GIS
View more work on GitHub →
Publication
ACM · 2025

Wildfire Detection From Landsat-8 Images Using Vision Transformer

Presented at the 6th International Conference on Computer Vision and Computational Intelligence (CVCI 2025) · Published in ACM Proceedings
doi.org/10.1145/3744725.3744728 →
Experience
Years across research fellowships, a GeoAI studio, and university innovation management.
Aug 2025 — Present
Graduate Assistant
Office of Innovation Management, UL Lafayette · Lafayette, LA
  • Maintain an IP database covering hundreds of patents and trademarks.
  • Built a web platform for querying and updating IP records, cutting lookup time 40%.
  • Research AI applications in IP commercialization, including patent analysis and market matching.
  • Support innovation and tech-transfer projects with researchers and entrepreneurs via the LEED Center.
Dec 2023 — Dec 2024
GeoAI Associate
NAXA Pvt. Ltd. · Kathmandu, Nepal
  • Processed and curated thousands of geospatial samples for analysis and modeling.
  • Built climate risk models from household, building, and socio-economic data.
  • Trained models across 10+ architectures — U-Net, InceptionNet, ResNet, VGG, YOLO, ViT, MedT, TransUNet, SAM.
  • Used LLMs and VLMs (Moondream2) for query automation, captioning, and object search.
  • Built ML pipelines and deployment for a platform serving trained geospatial models.
  • Led research collaboration and supervised junior teammates on ML and deployment work.
Sept 2023 — Dec 2023
GeoAI Fellow
NAXA Pvt. Ltd. · Kathmandu, Nepal
  • Researched problems in Nepal solvable with AI and aerial imagery.
  • Labeled and segmented 5,000+ geospatial samples for model training.
  • Trained U-Net, InceptionNet, ResNet, VGG, and YOLO across large-scale satellite imagery.
  • Collaborated with international mentors across 3 global projects.
  • Contributed to open-source tools for urban planning and damage assessment.
Jan 2023 — Dec 2023
AI Fellowship
FuseMachines · Kathmandu, Nepal
  • Completed 4 major ML/DL projects across supervised, unsupervised, and deep learning methods.
  • Completed a microdegree in NLP and Computer Vision.
  • Applied ML techniques to datasets ranging from 5K–50K samples.
Skills

Languages & Frameworks

Python C / C++ PyTorch TensorFlow Django FastAPI

Computer Vision & ML

OpenCV Scikit-Learn SAM / YOLO / ViT U-Net / ResNet / VGG Model Deployment · MLOps

Geospatial & Remote Sensing

QGIS ArcGIS Google Earth Engine Geopandas / rasterio LULC Mapping · Change Detection SAR & Optical Processing

LLM / NLP & Data

LangChain LlamaIndex Pandas / NumPy PostgreSQL · MySQL AWS · Docker Streamlit · Plotly
Education
University of Louisiana at Lafayette
M.S. in Computer Science · GPA 4.0/4.0
Expected Dec 2026
Nepal College of Information Technology
B.E. in Computer Engineering · GPA 3.81/4.0
Aug 2023

Training & Certification

AI · ML · Deep Learning — Fusemachines NLP & Computer Vision Microdegree — Fusemachines AWS Academy Cloud Foundation Pandas · Feature Engineering · Time Series — Kaggle Python for Everybody — University of Michigan (Coursera)