Leveraging Geospatial AI & PostGIS for Climate-Driven Disaster Mitigation
How spatial databases and machine learning can be combined to build predictive hazard maps that save lives — lessons from the NDMA ResilientPath AI project.
↗
NOORUL AMIN
AI & Robotics Undergraduate — Noor Ul Amin
"Standing still as the world evolves is our greatest risk—embracing artificial intelligence isn't optional, it is essential. AI won't replace human vision; rather, when we anchor intelligent systems in human purpose, we gain the foresight and speed needed to solve our hardest challenges and shape the world around us."
Building Intelligent Systems for a Resilient Planet.
I'm Noor Ul Amin, an AI & Robotics undergraduate at SZABIST, originally from Gilgit, Pakistan. I build real-time computer vision systems, geospatial disaster intelligence platforms, and interactive physics simulations — driven by a core mission: applying machine learning to environmental science and climate resilience.
Core Vision
Applying AI and ML to Environmental Science, Climate Resilience, and Disaster Mitigation. I believe the most urgent use of AI is protecting vulnerable ecosystems and the communities that depend on them.
Research Goals
Building AI-driven spatial models to analyze ecological degradation and water management. Combining computer vision with satellite imagery to build predictive environmental intelligence systems.
Education
B.Sc. Artificial Intelligence & Robotics — SZABIST, Islamabad. Expected June 2027.
HSSC in Computer Science — The Academy of Excellence, Zulfiqarabad, Gilgit (2022).
SSC in General Sciences — Elysian Higher Secondary School, Chinarbagh, Gilgit (2020).
Location & Contact
Based in Gilgit, Pakistan. Open to remote collaboration, research partnerships, and global internship opportunities.
Languages: English, Urdu, Punjabi · Basic Japanese & Spanish.
Community
Selected as one of 32 participants from 16 countries to represent Pakistan at the Global Encounters International Camp (Islamabad, Lahore & Karachi, Dec 2022 – Jan 2023). Led cross-cultural sessions promoting unity and diversity.
COMPUTER VISION — GEOSPATIAL AI — CLIMATE TECH — NDMA — YOLOV8 — DEEPSORT — POSTIS — FASTAPI — PYTORCH — COMPUTER VISION — GEOSPATIAL AI — CLIMATE TECH — NDMA — YOLOV8 — DEEPSORT — POSTIS — FASTAPI — PYTORCH —
Technical Stack
CORE SKILLS
AI & Machine Learning
Machine LearningDeep LearningNeural NetworksCNNRNNLSTMBiLSTMGRUVision Transformers (ViTs)Autoencoders
Computer Vision
OpenCVTensorFlowPyTorchYOLOv8DeepSORTImage ProcessingFace Recognition
Geospatial & Climate Tech
LeafletPostGISGIS Data MappingHazard AnalyticsSatellite ImagerySpatial Clustering
Programming Languages
PythonC++JavaJavaScriptRC#HTML / CSSAssembly Language
Backend & Web
FastAPINode.jsPWA (Progressive Web Apps)
Tools & Environments
Linux (Mint)Git / GitHubDockerWindows 11
WORK
001 / GEOSPATIAL AI / DISASTER TECH
DISASTERLENS AI
NDMA ResilientPath AI v2.0
Real-time disaster hazard aggregation and early warning geospatial platform built in 2025. Features automated spatial clustering, multi-hazard intelligence pipelines, and full offline PWA capability for field responders with zero connectivity.
YEAR: 2025
HAZARD
002 / COMPUTER VISION / TRACKING
MOTION-TRACKER
Real-Time CPU-Optimized MOT
High-performance multiple object tracking system combining YOLOv8 and DeepSORT, engineered for CPU-only environments. Delivers smooth, accurate tracking on lightweight hardware without requiring high-end GPU acceleration.
YEAR: 2025
TRACK
003 / PHYSICS SIMULATION
PENDULUM SIM
Multi-Pendulum Chaos Simulator
Interactive physics-based simulation modeling chaotic dynamical systems and multi-pendulum motion vectors. Demonstrates the butterfly effect through deterministic chaos using RK4 numerical integration.
YEAR: 2025
CHAOS
004 / ASSISTIVE AI / DEEP LEARNING
ASSISTIVE CV
Assistive Application for Visually Impaired
CNN-based real-time face recognition and YOLO object detection system designed for visually impaired individuals. Combines robust image pre-processing pipelines with live camera feeds for ambient spatial awareness and independent navigation.
YEAR: 2024
VISION
005 / GAME DEVELOPMENT / C#
BOUNCE 2026
Interactive 2D Arcade Physics Game
Custom interactive 2D arcade physics game with precise collision detection and smooth gameplay mechanics, deployed to the Microsoft Store platform for public distribution.
YEAR: 2026
ARCADE
Professional Experience
WHERE I'VE WORKED
July 2025
AI & Digital Literacy Mentor
ITREB — Crossroad Camp, Hundur, Pakistan
- Facilitated interactive sessions on digital literacy for 13–15-year-old youth participants.
- Introduced core concepts of Artificial Intelligence, sparking interest in emerging technologies among young learners.
- Guided discussions on global citizenship to promote ethical awareness and pluralism.
- Designed hands-on activities blending AI tools with real-world problem-solving skills.
- Mentored small groups to build confidence in using digital resources responsibly.
2025
DisasterLens AI — NDMA Pakistan Project
National Disaster Management Authority (NDMA), Pakistan
- Engineered NDMA ResilientPath AI v2.0 / DisasterLens AI — a real-time hazard aggregation and early warning geospatial platform.
- Built offline-first Progressive Web Apps (PWAs) enabling field responders to access critical maps in zero-connectivity environments.
- Automated multi-hazard intelligence pipelines and integrated spatial clustering (DBSCAN) with Leaflet & PostGIS for live disaster mapping dashboards.
July 2024
English Language Tutor
Native Language Camp — International Students
- Tutored international students in English, enhancing language fluency and overall academic confidence.
Research & Writing
TECHNICAL BLOG
Optimizing Real-Time Multi-Object Tracking (YOLOv8 + DeepSORT) on CPU
A deep dive into the engineering trade-offs required to run accurate, high-FPS object tracking on commodity hardware — no GPU required.
↗Simulating Chaotic Physics Systems in C++
Why the multi-pendulum is the perfect gateway into deterministic chaos, and how to build a high-fidelity simulator that reveals the butterfly effect in action.
↗