John V. Kenny — Data & Spatial Intelligence

About

I build systems that transform raw spatial data into civic intelligence — automated pipelines, interactive dashboards, and computer vision workflows that help municipalities operate smarter.

Background

With a Master's in Political Science and years spent leading mission-driven, people-focused programs, I bring a strategic, research-driven mindset to every technical challenge — bridging the gap between complex data systems and the people who depend on them.

What I Build

Four things I keep
coming back to

Document & Image Intelligence

Turning unstructured records — scanned drawings, handwritten cards, imagery — into structured, queryable data, with a human in the loop before anything becomes authoritative.

Data Quality at Scale

Catching the structural problems that break geocoding, routing, and billing — systematically and before delivery, rather than after something fails in the field.

Public-Facing Civic Apps

Tools residents actually use, built for the phone in someone’s hand — and designed so privacy and transparency don’t have to trade against each other.

Spatial Systems & Automation

The pipelines, databases, and cloud plumbing underneath — where a manual process becomes a scheduled one that nobody has to remember to run.

Selected work

01

Featured • The News Atlas

The news, mapped like an atlas

1 / 5

Map of Brazil's 5,571 municipalities coloured by the 4 October 2026 presidential first round: rust for Lula, teal for Bolsonaro, darker for bigger margins, with the Northeast almost solid rust. An inset sizes each municipality's lead in votes, a bar chart gives net votes by region, and a grid of figures shows all 158.7 million voters. — News Atlas

Two Brazils

Brazil, 4 October 2026. The presidential first round by municipality, from the electoral court's polling-station results. Full resolution →

Map of central Panama after the magnitude 7.7 Azuero earthquake: modelled shaking, the fault rupture, aftershocks and a tsunami-warning inset. — News Atlas

The Azuero Earthquake

Panama, 9 October 2026. USGS shaking, rupture and aftershocks, mapped four days after the shock. Full resolution →

Map of three October 2026 hurricane tracks: Isaias through the Gulf to the Florida Panhandle, Simon up Mexico's Pacific coast, and Rachel looping out into the Pacific and back into Baja California, with wind swaths, a wind-speed chart and numbered notes. — News Atlas

Three Hurricanes, One Weekend

North America, 9–11 October 2026. National Hurricane Center best tracks; line width is wind speed. Full resolution →

Two maps of the Midwest side by side: Monday 12 October's record highs in orange-brown, with badges at the 20 stations that set records, and Saturday 17 October's forecast lows in blues, below freezing across the north. A chart tracks Chicago's October highs and lows. — News Atlas

From Record Heat to Frost

The Midwest, 12–17 October 2026. Monday's record highs from NOAA's hourly analysis beside Saturday's forecast frost. Full resolution →

Poster map of the Chicago region: lidar relief, the shorelines of glacial Lake Chicago, and historic trail roads colored by how far they stand above their flanks. — Study · with Claude Fable 5.1

The Ghost Shorelines of Chicago

A lidar test of whether Chicago’s diagonal streets ride the beach ridges of a glacial lake. Half of them do. Watch the film & see the map →

All the maps & commissions →

02

AI-Powered Infrastructure Intelligence

Computer Vision VLMs Python Azure AI ArcGIS Pro

Problem, approach & result

Problem

Municipal governments sit on decades of as-built drawings, handwritten service records, and construction documents that are trapped in paper or unstructured PDFs — invisible to the GIS systems staff rely on every day.

Approach

A suite of computer vision and AI pipelines that extract structured data from these documents at scale — from vision language models that read handwritten water service cards to automated as-built ingestion that pulls metadata, organizes files, and feeds results directly into ArcGIS Pro catalogs, with a human-approval gate before anything enters the catalog of record.

Result

Decades of paper became a living, queryable system — from scanned drawing, to georeferenced GIS feature, to natural-language answer. Higher-quality infrastructure data with traceable lineage back to the source record. Building it stretched me across computer vision, model training, cloud architecture, agentic AI, and RAG.

03

Downtown Parking App with Walk-Time Routing

React ArcGIS Maps SDK Walk-Time Routing Custom Web App

Problem, approach & result

Problem

A lakefront suburb needed residents and visitors to understand downtown parking at a glance — including “how far is this lot from where I'm going?” — and off-the-shelf ArcGIS Experience Builder couldn't deliver that experience.

Approach

I designed, built, and deployed a custom React + ArcGIS app: every CBD lot mapped with color-coded restrictions and detailed profiles, on-demand walk-time routing from any origin, and a mobile-first interface — all driven by a single JSON config that makes it portable to other communities.

Result

I owned it end to end — requirements, stakeholder data validation, build, and deployment — with no internal template to work from. Because I built it config-driven, a second municipality was served from the same codebase with no code changes.

Read the case study →

See all work by capability →

Additional Projects

More from the workshop

GeoAI

Spatial data and modern models, on my own time. A vision LLM that identifies a plant or insect from a phone photo and files it by place and date. A location engine that turns years of raw GPS into stay-points, mode-classified trips, and street-matched drive tracks. An atlas that finds the eras of a life by detecting when home moved.
Leah Rae Genealogy

The same document-intelligence capability on my own time: AI agents over Azure Postgres/PostGIS read ~1,400 pages of historical records, transcribe them, and align every fact to the right ancestor with full citations. Broke a 150-year research wall — and corrected an error in the authoritative source.
Visit →
First Return

Turns real USGS lidar into 3D-printable models of actual places — a small product studio built on real elevation data.
Visit →
SettleStack

Full-stack poker debt settlement app — calculates optimal payment paths to minimize total transactions among players.
Utility Document Extractor

Document-intelligence pipeline that parses inconsistent utility as-built documents and structures the data for GIS integration — eliminating slow, error-prone manual entry.
Utility Atlas Requests Automation

End-to-end Python + Power Automate workflow automating atlas request intake, spatial processing, and document delivery — cutting per-request turnaround from ~8-10 min to ~1-2.
Emergency Vehicle Preemption Dashboard

Real-time ArcGIS dashboard consolidating preemption device status and health into a single operational view for emergency response coordination.
Historical Records Toolkit

A reusable, config-driven system for digitizing scanned historical records with a vision-LLM and human-in-the-loop review. Grew from digitizing several thousand handwritten water service cards — an estimated 40+ hours of manual entry, completed in roughly 4 hours of review.
Police Transparency Dashboard

Python pipeline that geocodes, categorizes, and anonymizes incident data to the 100-block level, then auto-publishes a public dashboard — 60-90 minutes of manual work down to about 5.
Visit →

Expertise

Core competencies

GIS & Spatial Analysis: ArcGIS Pro, ArcGIS Online, spatial data pipelines, geodatabases, cartographic design, and field data collection systems.
AI & Computer Vision: RT-DETR object detection, Qwen2.5-VL and vision-language models, model training, automated feature extraction, and intelligent document processing.
Agentic AI & MCP: Building AI agents and Model Context Protocol (MCP) servers, with Claude Code as a daily development driver — including agentic architectures integrated with Microsoft 365 and Copilot Studio.
Cloud & Deployment: Azure Container Apps, Static Web Apps, Blob Storage, and PostgreSQL/PostGIS — deploying cloud-native, RAG-enabled applications and job pipelines.
Python & Automation: Python scripting, Power Automate, ArcPy, workflow automation, API integration, and process optimization.
Data Visualization: Interactive dashboards, ArcGIS StoryMaps, data-driven cartography, and dynamic reporting solutions.
Research & Communication: Master's-level research methodology, technical writing, stakeholder engagement, and cross-functional collaboration.
Project Management: End-to-end project delivery, requirements analysis, agile workflows, and multi-stakeholder coordination.

Let's build something together

Get in Touch View All Work

John V.Kenny