Small Batch Maps

Small Batch Maps

Fast, transparent geospatial tooling for climate, public health, and conservation teams.

My Services & Expertise

AI-enhanced spatial workflows, guided by care & domain expertise

Geospatial Data Science

I don't just clean data—I interrogate it. With Python, geopandas, and a healthy skepticism for null-filled shapefiles, I build workflows that turn chaos into clarity. Pattern-finding, truth-surfacing, and always auditable.

Web Cartography

A good map doesn't just inform—it moves. I design web maps and spatial dashboards that load fast, look sharp, and do exactly what they should. From climate stories to conservation tools, every pixel has a purpose.

AI-Assisted Workflows

I fold large language models and intelligent agents into GIS workflows without losing the plot. Think: faster code, automated insights, and spatial systems that scale—without falling for hype or one-click magic.

Exploratory Geospatial Data Science (ESDA) brings together location intelligence and exploratory data analysis to uncover patterns, trends, and relationships in our built and natural worlds. I help clients turn messy, siloed datasets into clean, interoperable formats and curated insights—ready to inform smarter decisions in planning, policy, logistics, public health, and client-specific needs.

My Approach to Spatial Science

Raw Geographic Data

Spatial datasets, APIs, field collection inputs

Data Cleaning & Analysis

Python + Jupyter for wrangling, exploratory analysis, and QA/QC

Visual Insights

Clear charts, spatial stats, and storytelling visuals

Scalable Infrastructure

Custom APIs, spatial databases, and reproducible pipelines

Interactive Web Maps

Custom UX/UI for intuitive data exploration

Core Capabilities

Exploratory Spatial Data Analysis (ESDA)

Identify spatial clusters, trends, and outliers using robust statistical and visual methods.

AI-Assisted Predictive Modeling

Forecast future scenarios using spatial machine learning and synthetic data generation.

Time-Series & Spatiotemporal Modeling

Understand how geographic phenomena change over time — seasonal patterns, mobility, and dynamic shifts.

Data Integration & API Design

Clean, combine, and operationalize datasets via modern spatial databases and scalable API pipelines.

I use LLMs to speed test writing, style generation, and refactors—inside guardrails. You get cleaner code, better docs, and fewer regressions. No model touches your data without approval, and every suggestion ships behind tests.

Refactor brittle GeoPandas script

Split 1k lines into testable modules; remove hotspots

Repo →

Generate MapLibre styles

Turn palette briefs into JSON styles; iterate fast

Repo →

Write failing tests

Surface edge cases in hydrology/ETL pipelines

Repo →

Explain this notebook

Extract decisions/assumptions into readable docs

Repo →

Catch logic gaps in ETL

Diagram flows; propose safer control paths

Repo →

TileCraft application interface showing AI-assisted vector tile generation

A Case Study in Human-AI Collaboration: Tilecraft

Tilecraft is an AI-assisted CLI tool I built that demonstrates these principles in action. It combines OpenStreetMap data processing with AI-powered schema generation and prompt-based style design. Users provide natural language descriptions like "subalpine dusk" or "urban midnight" and the system generates complete vector tile pipelines with beautiful, context-aware cartographic styles.

View on GitHub

Featured Deep-Dive

Building Production AI-Assisted Software

A comprehensive 5-part technical case study covering architecture patterns, testing strategies, and lessons learned from building Tilecraft—from prototype to production.

Architecture Patterns Testing Strategies Performance Analysis

Read the Complete Case Study

                                                
                                                # AI-assisted schema generation

                                                def generate_schema(prompt):

                                                  "subalpine dusk" →

                                                  complete vector pipeline

Recent Projects

Evidence over adjectives. A few representative builds and experiments that show how I work and what changes for a user.

WHO Outbreak Dashboard

Situational awareness for outbreaks—clear visuals for faster briefings

JS Dashboard Public Health

Live

Satellite Embeddings Change

Visualize spatial change via embedding differences

JS Embeddings Visualization

Repo

FlowFinder

Compare watershed boundaries across tools for accuracy and consistency

Python Hydrology ESDA

Repo

Cartographic Style Editor

Interactive style design for Colorado infrastructure

JS MapLibre Front-end

Repo

MBTiles Viewer

Explore and inspect vector tile datasets interactively

JS MapLibre Vector Tiles

Repo

Ghost Forest Watcher

Post-fire recovery monitoring with Sentinel-2 indices and lightweight ML

Python Sentinel-2 ML

Repo

About Richard Donohue

Crafting Maps with Purpose

With a Ph.D. in Geography, GIS, and Cartography, I specialize in transforming complex geospatial data into meaningful insights and beautiful, functional maps. My approach combines traditional cartographic principles with modern web technologies and current AI-assisted workflows.

I believe in ethical data practices and human-centered design. Every project is treated as a unique challenge, deserving custom solutions rather than cookie-cutter approaches. Whether you need sophisticated spatial analysis, interactive web applications, or AI-assisted workflows, I bring both technical expertise and creative vision to your geospatial challenges.

Born and raised in Durango, CO, I enjoy the outdoors, hiking, and skiing.

Open to full-time and contract roles — remote or Mountain West.

University of Kentucky — Program Architect & Assistant Professor (2014–2024)

Challenge: Build a cutting-edge graduate program in digital cartography that met professional standards and stayed reproducible.

What I built: A 9-course online MS/certificate program in Digital Mapping, spanning web mapping, geospatial analysis, UX design, and open-source development. Designed all core curriculum, championed reproducible workflows using Jupyter + version control.

Impact: 30+ MS and 50+ certificate graduates placed at USGS, Cornell, and The New York Times. Student projects won national NACIS and CaGIS awards.

Evidence: Program page, Student work gallery

Ready for collaboration?

Get in Touch

Name *

Email *

Comment *

I typically respond within 24 hours. Most projects start with a free discovery call—to find the signal in the noise and chart a direction together.

Let's Make Something Thoughtful Together

Whether you're mapping recovery networks, building smarter dashboards, or just want to talk shop about spatial ethics—I'm all ears.

My Services & Expertise

Geospatial Data Science

Web Cartography

AI-Assisted Workflows

My Approach to Spatial Science

Raw Geographic Data

Data Cleaning & Analysis

Visual Insights

Scalable Infrastructure

Interactive Web Maps

Core Capabilities

Exploratory Spatial Data Analysis (ESDA)

AI-Assisted Predictive Modeling

Time-Series & Spatiotemporal Modeling

Data Integration & API Design

Refactor brittle GeoPandas script

Generate MapLibre styles

Write failing tests

Explain this notebook

Catch logic gaps in ETL

A Case Study in Human-AI Collaboration: Tilecraft

Building Production AI-Assisted Software

Recent Projects

WHO Outbreak Dashboard

Satellite Embeddings Change

FlowFinder

Cartographic Style Editor

MBTiles Viewer

Ghost Forest Watcher

About Richard Donohue

Crafting Maps with Purpose

Experience at a Glance

University of Kentucky

World Health Organization

Ready for collaboration?

Get in Touch

Let's Make Something Thoughtful Together