Dataviz

D3 experiments, Observable notebooks, and visualization techniques

Broadcast Television

NBC News / MSNBC Big Board

Senior product designer on the NBC News Digital team for the groundbreaking "Big Board" election visualization system used by Steve Kornacki on MSNBC and NBC News . The system launched for the 2018 midterm elections.

Technical Architecture:

Hardware: Gaming PC (Intel i7 Quad Core, ATI Radeon GPU) with Planar touchscreen displays
Software: HTML5 web app in Chrome on Windows 10 kiosk mode
Libraries: D3.js for visualizations and animations, Vue.js for transitions
Data: Live election results with historical context (10+ years)

Features:

Kornacki Scorecard - Visualization of 60+ critical races
What-if scenarios - Real-time vote modeling during broadcasts
Data Cross-Filter - Demographic mapping and voting pattern analysis (created with Shirley Wu)
Granular mapping - Zoom to individual districts with historical data

Project collaborators: Michael Small, Two-N, Shirley Wu

Observable Notebooks

Personal experiments and data visualizations: observablehq.com/@ejfox

Election & Political Data

Wikidata Election Data / Candidate Exploration - Query builder for election data analysis
Why and How to Focus on Outstanding Ballots - 2020 election analysis methodology

COVID-19 Visualizations

CMU COVIDcast API: Bubbles Textures - Force-directed county visualization with population-based circles
Population Contour Visualization - Population density mapping experiments

Experimental & Creative

I Ching Hexagrams - Traditional divination system visualization
12 Name Clock - Custom clock interface experiment
Random Ingredient Generator - Creative recipe generation tool

Data Integration

Pinboard Bookmarks Integration - Personal data visualization techniques

News Industry Projects

Vocativ (2014-2016)

Joined as Graphics Editor to build data visualization team from scratch. Scaled production from 5-6 graphics per month to 5-6 per day through process optimization and team building .

Daily Graphics Production:

Hundreds of infographics created by hand
Breaking news response systems
Style guides and production templates
Team training and workflow optimization

Dataproofer

Knight Prototype Fund project: "spellcheck for data" - automated dataset error detection tool

Project Details:

Funding: Knight Foundation Prototype Fund
Repository: github.com/dataproofer/Dataproofer
Distribution: dataproofer.org • Electron app • NPM package

Technical Implementation:

Test Suites: Core, Info, Statistical, Geographic
Input Formats: CSV files, published Google Sheets
Output: JSON results, human-readable reports
Architecture: Modular test system with lodash utilities

Test Categories:

Core Suite - Common data problems (truncation, formatting)
Info Suite - Descriptive statistics and data profiling
Statistical Suite - Outlier detection and distribution analysis
Geographic Suite - Location data validation

Freelance & Early Work

Visual.ly (2012-2014)

Staff Designer at infographic startup Visual.ly in San Francisco. Learned D3.js and JavaScript from incredibly smart coworkers .

Independent Freelance

Hundreds of custom infographics for startups, VC presentations, and business explanations .

Technical Approach

Core Technologies

D3.js - Primary data visualization library
JavaScript/ES6 - Modern development practices
Observable - Reactive notebook environment for prototyping
Vue.js - Component-based UI framework

Data Sources

Election APIs - Real-time voting data and historical records
Government datasets - Census, demographic, geographic data
Health data - COVID-19 tracking and epidemiological datasets
Personal APIs - Custom data integration and tracking

Creative Experiments

Generative & Algorithmic

Mathematical visualization experiments
Procedural content generation
Interactive data exploration tools
Esoteric data visualization (I Ching, random generators)

Last updated: June 2025