Skip to main content

Reference

Knowledge Base

Short-form, evergreen explainers on the concepts, tools, and systems that underpin my research — from geospatial foundation models and hexagonal grids to cloud-native data infrastructure and conflict data engineering.

7 articles  ·  Updated regularly

Tools & Libraries · Intermediate

Cloud-Native Geospatial Architecture

Cloud-native geospatial architecture is an approach to building geographic data systems using cloud-first design principles: object storage as the primary data store, columnar formats for analytics, serverless compute for ETL, and API-first access.

4 min read
Tools & Libraries · Intermediate

Working with OpenStreetMap at Scale

OpenStreetMap (OSM) is the world's largest open geographic database. Processing its full planet file for machine learning and spatial analytics requires specialised parsing strategies, efficient data structures, and cloud-scale infrastructure.

4 min read
GeoAI · Intermediate

Place Embeddings & Spatial Representation Learning

Place embeddings are dense vector representations of geographic locations learned from spatial data. They encode a location's semantic character — its land use, neighbourhood context, and functional role — in a form usable by machine learning models.

4 min read
Spatial Data · Foundational

Discrete Global Grid Systems (DGGS)

A Discrete Global Grid System (DGGS) is a spatial reference framework that partitions the entire surface of the Earth into a hierarchical tessellation of equal-area, gapless cells — enabling consistent multi-scale geographic analysis.

4 min read
Spatial Data · Foundational

H3 Hexagonal Grid System

H3 is Uber's open-source hexagonal hierarchical geospatial indexing system. It divides the Earth into a multi-resolution grid of hexagonal cells, providing consistent spatial indexing for large-scale location analytics.

4 min read
GeoAI · Intermediate

Graph Neural Networks for Spatial Analysis

Graph neural networks (GNNs) are a class of deep learning model designed for graph-structured data. In spatial analysis, they learn representations of places, roads, and regions by propagating information across geographic adjacency graphs.

4 min read
GeoAI · Intermediate

Geospatial Foundation Models

Geospatial foundation models are large pre-trained neural networks adapted to understand and reason about spatial data — from satellite imagery and map tiles to GPS traces and place embeddings.

3 min read