duncan richey

Hi, I'm Duncan.

I'm a transportation planner and data analyst who kept ending up as the person who just builds the thing instead of asking someone else to. I work in the space between product, maps, and code, and with AI in the loop I ship a lot faster than that used to be possible.

project spotlight

Tahoe Region Evacuation Planning Tool Tahoe Region Evacuation Planning Tool An interactive sandbox for modeling wildfire evacuation demand across the Lake Tahoe region. Townsquare Townsquare A mobile app that runs my Blood on the Clocktower game nights, from invites to results.

transportation · data · spatial tech

things i build for cities

A few years of moving crashes down, dragging housing rules into the open, and getting street redesigns onto the ground, mostly by turning messy public data into something people can actually use.

Tahoe evacuation tool placeholder

Tahoe Region Evacuation Planning Tool

An interactive demand modeling tool for regional wildfire evacuation scenarios around Lake Tahoe.

Solo build
Microsoft Power BI logo

Safe Streets for All: data analysis at VHB

Crash analysis and interactive dashboards for 15+ Safety Action Plans across North Carolina, Colorado, and New York.

2023–2024 · Data analyst
New York City Department of Transportation logo

NYC DOT: Vision Zero project management

A summer redesigning two NYC streets, one corridor in Astoria and one intersection in the Bronx, for the team trying to get serious-injury crashes to zero.

2023 · Summer intern, Research & Implementation Services
NC Zoning Atlas map view

North Carolina zoning atlas

A map of 389 zoning districts across Raleigh and Cary, built from 1,500 pages of ordinance text, to make local housing rules legible.

2023 · Master's project lead
Line drawing of an electric bicycle

Durham Bull E-Bike Pilot: project management

An e-bike pilot for 61 downtown Durham hospitality workers, and the evaluation my UNC team ran to find out whether it got them out of their cars.

2022–2023 · Project Manager

nights, weekends & communities i love

things i make for people

The other half of how I work: small, fast, and for people I actually know. A nonprofit, a watch party, a birthday gift. It's where I prove I can take something from idea to shipped over a weekend.

The Townsquare landing page

Townsquare

A mobile-first web app for managing private Blood on the Clocktower game night communities.

Solo build (in progress)
SAVANNIMALS

SAVANNIMALS

A Survivor fantasy draft app I built and run for a 31-player league of my friends, rebuilt from a Google Sheet into a full-stack app.

2024 · Solo build (in progress)
Flordle title card

Flordle

A custom Wordle clone, built one daily puzzle at a time, as a birthday gift.

2023 · Solo build
A stylized credit card

Spend a CEO's salary

An interactive site that turns C-suite pay into things you could buy with it instead.

2024 · Solo build (in progress)
Carolina Thrift logo in a multicolored circle

Carolina Thrift

A student-run nonprofit at UNC that turned dorm move-out trash into a 3,000-person sale. I founded it.

2018–2021 · Founder
Carolina Kickoff banner

Carolina Kickoff

Co-directed a three-day, $40K orientation camp at UNC for 200 incoming first-years.

2019 · Co-director

on film

photography

I carry a point-and-shoot film camera most places. Prints are at darkroom.com.

Duncan Richey

about

a little more about me

I started out as an urban planner, the kind who's happiest with a map open and a question that doesn't have a clean answer yet. Over the last few years that turned into a data career: crash analysis at VHB, a summer redesigning streets at NYC DOT, a zoning atlas that took 1,500 pages of ordinance and made it legible. Somewhere in there I noticed I'd become the person on every team who just builds the tool instead of waiting for one, and with AI in the loop I can now go from idea to working thing faster than ever.

I'm looking for a product or forward-deployed role at a company working on transportation, cities, or spatial data, somewhere I can sit close to real users and ship. Outside of work I run a thrift nonprofit, a Survivor watch party, and a silent-disco headphone business, which is to say: I like making things for people, and I don't really turn it off.

The best way to reach me is at duncan527@gmail.com.

things i make for people

Townsquare

A mobile-first web app for managing private Blood on the Clocktower game night communities.

Solo build (in progress)
Open the app ↗

Townsquare is a mobile-first web app for managing private Blood on the Clocktower game night communities. It replaces the manual workflow of spreadsheet tracking and individual SMS coordination with a streamlined system for invites, RSVPs, queue management, and post-game recording, anchored by a phone-and-SMS-code auth flow.

The first community on the platform is my own where I host around four sessions a month for a pool of ~70 friends. Blood on the Clocktower is a social deception game akin to Mafia or One Night Werewolf that I have become obsessed with over the last year. Texting 20+ people each game, tracking results, and managing players became too time-intensive and annoying for me. So, with the help of Claude Code, I’ve been developing this since May 2026. I’m building it solo on Next.js, Supabase, Twilio, and Vercel, with multi-tenancy on the roadmap so other communities can run their own.

The Quincywood Bluff game night page showing claimed and open seats
A player profile with games played, win rate, and past game history
The post-game recap showing every role in play and which team won
Claiming a seat for game night, a player's record, and the recap once the night is over.

things i make for people

Carolina Kickoff

Co-directed a three-day, $40K orientation camp at UNC for 200 incoming first-years.

2019 Co-director
Website ↗Press: Daily Tar Heel ↗Instagram ↗

Carolina Kickoff is a three-day, student-run camp that throws 200 incoming first-years together right before they start at UNC, so they show up to college already having a few people who feel like home. I was a counselor for two years and a co-director my third year.

That meant running a $40,000 production with a team of 40 students: picking and training the counselors and committee chairs, wrangling enrollment for 175 campers, and keeping three days of programming from falling apart. The thing I’m proudest of is the money. We raised an extra $10,000 specifically so that, for the first time, anyone who got in could come for free regardless of whether they could pay. We also spent real time questioning which long-standing camp traditions actually made everyone feel welcome, and changed the ones that didn’t.

It was a crash course in the unglamorous half of building anything: budgets, people, logistics, and the calls you make when something’s about to go sideways at 11pm.

Carolina Kickoff
1 / 2

things i make for people

Carolina Thrift

A student-run nonprofit at UNC that turned dorm move-out trash into a 3,000-person sale. I founded it.

2018–2021 Founder
Website ↗Press: Daily Tar Heel ↗Instagram ↗

Every spring, college students throw out an absurd amount of perfectly good stuff on move-out day: futons, mini-fridges, lamps, clothes still with tags. I started Carolina Thrift to catch all of it.

The idea was simple and the logistics were not. Collect the move-out castoffs from the dorms, store it all over the summer, and sell it back in one giant fall sale. I won a campus pitch competition for the seed money, pulled together a team of fourteen students, and we threw the first sale to a crowd of over 3,000 people. In my years running it we raised about $20,000 and kept 20 tons of stuff out of the landfill.

The part that still gets me is that it outlived me. Carolina Thrift is a real 501(c)(3) now, has raised more than $50,000, funds student sustainability projects, and is still running every year. I got to talk about how it started at the 2019 Sustainable Brands conference. It was the first big thing I built that turned into something other people kept building, which, it turns out, is the kind of work I keep chasing.

Carolina Thrift
1 / 2

things i make for people

Spend a CEO's salary

An interactive site that turns C-suite pay into things you could buy with it instead.

2024 Solo build (in progress)

This one started as a feeling: the numbers in CEO pay disclosures are so big they stop meaning anything. So I’m building a site that makes one of them mean something. You pick a real CEO, and then you “spend” their annual pay on things you can actually picture, like a $10,000 raise for every person at the company or a few miles of protected bike lane, while the receipt fills up in real time.

I designed the credit-card components and the layout in Figma and I’m coding it in HTML, CSS, and JavaScript, with animated cards that track your spending as you go. It’s still in progress: equal parts a way to sharpen my front-end and a small argument about inequality that you get to make with your own hands instead of just reading.

things i build for cities

Durham Bull E-Bike Pilot: project management

An e-bike pilot for 61 downtown Durham hospitality workers, and the evaluation my UNC team ran to find out whether it got them out of their cars.

2022–2023 Project Manager
About the program ↗ Executive summary (PDF) ↗Implementation guide (PDF) ↗

If you hand someone an e-bike, do they actually stop driving? The City of Durham decided to find out, and my team at UNC got to be the people who answered.

The Bull E-Bike Pilot came out of a $1 million Bloomberg Mayors Challenge grant Durham won in 2018 to cut down on people driving downtown alone. What made it unusual is who it went after. Instead of recruiting the cyclists who would have shown up anyway, the city aimed at downtown hospitality workers, the restaurant and brewery and hotel staff whose shifts end after the buses stop. Preference went to applicants who were non-white, not cis men, over 45, non-English speaking, lower income, and transportation-disadvantaged. 322 people applied and 61 got bikes, across seven cohorts between August and November of 2022. Each one came with a helmet fitted to the person’s actual head, a lock, and somewhere to take the bike when it broke. Of the people who rode, 74% were non-white, the median household income was $40,000, and 40% were working two or more jobs.

The city contracted our second-year master’s cohort at UNC’s Department of City and Regional Planning to evaluate the whole thing, and I ran that engagement as project manager. Our half was the evidence: aggregating the anonymized trip data coming out of OpenPATH, the tracking app NREL built for the pilot, measuring it against the program’s key performance indicators, sitting down for one-on-one exit interviews with participants as each cohort finished, and writing the final report of findings and recommendations that went back to the city.

The numbers held up. Participants logged more than 1,100 confirmed e-bike trips, over a third of every trip they took during the pilot, and rode 2,933 miles. More than half of those trips (54.6%) were commutes. Most importantly for a program built to reduce drive-alone travel: 38.1% of e-bike trips replaced a trip someone would otherwise have driven alone, which worked out to 1,099 miles of solo car travel that simply didn’t happen. The qualitative side told the part the app couldn’t. People without reliable access to a car changed their travel the most, and one participant did the math on her own gas bill and figured she’d saved $150 to $200 in a month.

The program went on to win the Association for Commuter Transportation’s Outstanding Diversity, Equity and Inclusion Award and an Equity in Cleantech Innovation Award in 2023, plus Special Recognition at the Governor’s Awards for High-Impact Research. Durham published an implementation guide so other cities could copy it, and the evaluation we wrote is a chunk of why there was anything worth copying.

What stuck with me is how much of the result came from decisions made long before any data existed. The pilot worked because someone chose to recruit line cooks instead of commuters who already owned Stravas, and because the city was willing to be measured honestly on whether that worked. My job was to make sure the answer was trustworthy either way.

things i make for people

Flordle

A custom Wordle clone, built one daily puzzle at a time, as a birthday gift.

2023 Solo build
Play Flordle ↗

I taught myself HTML, CSS, and plain JavaScript by building my girlfriend (now wife, Flo) a birthday present. It’s a Wordle clone, a new “Flordle” every day for a week, but the puzzles were rigged. Certain letters from each day’s word were hers to collect, and by the end of the week they unscrambled into “Sweeney Todd.” That was the real gift: tickets to the show.

It’s the first thing I ever coded from nothing, and it’s still my favorite reason to learn a new tool. Make something for one specific person you love, and don’t stop until it works.

Flordle
Flordle
Flordle
1 / 3

things i build for cities

North Carolina zoning atlas

A map of 389 zoning districts across Raleigh and Cary, built from 1,500 pages of ordinance text, to make local housing rules legible.

2023 Master's project lead
Open the atlas ↗ Final paper (PDF) ↗

Zoning is the quiet thing that decides what can get built where you live, and it’s written across 1,500 pages of ordinance that almost nobody reads. For my Master’s project I tried to fix the readability problem the only way I could think of: I read all 1,500 pages and turned them into a map.

The finished atlas covers 389 zoning districts across Raleigh and Cary, the two big neighbors in North Carolina’s Triangle, home to about 650,000 people. Once the rules are a map instead of a legal document, the differences jump out: how much more freely one city allows multi-family housing than the other, or where an accessory dwelling unit (a backyard cottage, basically) is and isn’t legal. That’s not trivia. It’s a big part of why housing costs what it costs.

The data fed into the North Carolina Zoning Atlas and the broader National Zoning Atlas, which is the part I’m proudest of. It stopped being my class project and became something advocates and legislators could actually use. You can open the live map below, or read the full paper if you want the methodology.

North Carolina zoning atlas
North Carolina zoning atlas
1 / 2

things i build for cities

NYC DOT: Vision Zero project management

A summer redesigning two NYC streets, one corridor in Astoria and one intersection in the Bronx, for the team trying to get serious-injury crashes to zero.

2023 Summer intern, Research & Implementation Services

During the summer after my first year of graduate school, I worked with NYC DOT’s RIS team— a division focused on Vision Zero and eliminating serious injuries on city roads— where I contributed to multiple street improvement projects to enhance safety for all road users by collecting and analyzing crash and traffic data, mapping project locations, conducting field work, and supporting policy research to inform design and community engagement.

I spent most of my time on two re-designs. One was a two-mile corridor in Astoria, Queens, and the other was a large three-way intersection in the Bronx that had become a spot people came to street-race. The before-and-after of that intersection is below. Collaborating with planners and engineers, we narrowed crossing distances, repositioned stop lines, and introduced large, painted curb extensions, effectively “right-sizing” the intersection to discourage unsafe maneuvers. These interventions improved pedestrian access, traffic operations, and overall road safety.

The part I loved most was that none of it was guesswork. Every move started in the crash and traffic data, got checked against what I actually saw standing on the corner doing fieldwork, and ended up as a preliminary design I drew in AutoCAD. It was the first time I felt the whole loop: data, to a decision, to a thing that changes how a real street works while navigating the department and political realities that move a city.

The same intersection after right-sizing The Bronx intersection before the redesign

things i make for people

SAVANNIMALS

A Survivor fantasy draft app I built and run for a 31-player league of my friends, rebuilt from a Google Sheet into a full-stack app.

2024 Solo build (in progress)
SAVANIMALS ↗Come On In Gays ↗

SAVANNIMALS is a Survivor fantasy draft web app I built and administer for a 31-player league of my friends that runs each season alongside the show. The original version, built on Google Apps Script and Google Sheets primarily with Gemini, handled drafts, scoring, and leaderboards, but every episode required a tedious round of manual updates.

I’m currently using Claude Code to rebuild the current web app on a modern full-stack foundation (Next.js, Supabase, Drizzle, shadcn/ui, Vercel, and Twilio for SMS auth), with the long-term goal of fully automating the post-episode workflow via a Survivor Wiki scraper and replacing the original system’s redundant calculation sheets with live SQL queries.

Come On In Gays is a Survivor watch party I help run here in Brooklyn. As part of my AV job for the show, I’ve been piloting small web apps to facilitate more audience interaction and encourage repeat visits. I am in the process of expanding the webapp to incorporate a second league for our Come On In Gays audience.

Pictured is the previous webapp, live via Google Apps Scripts.

The SAVANNIMALS season leaderboard
A player's tribe dashboard with season scoring
The media and stats feed
The Apps Script version, as the league used it on their phones.

things i build for cities

Tahoe Region Evacuation Planning Tool

An interactive demand modeling tool for regional wildfire evacuation scenarios around Lake Tahoe.

Solo build
Open the live tool ↗

The Lake Tahoe Wildfire Evacuation Sandbox is an interactive demand modeling tool for regional wildfire evacuation scenarios. It addresses a real gap in the official 185-page Lake Tahoe Regional Evacuation Plan, which uses entirely qualitative route selection criteria and has no mechanism for the summer tourist surge that roughly doubles the area’s effective population.

I built the MVP in React, Vite, and Mapbox GL, with segment-level corridor scoring in Turf.js against CAL FIRE hazard severity zones, Replica mobility data, and ACS demographics. Users can toggle hazard scenarios and see how present-population estimates and vehicle conversion factors shift which corridors get prioritized.

The project started as my response to a CalTrans case interview prompt and became a working argument for what a modern, data-driven evacuation planning interface could look like.

things i build for cities

Safe Streets for All: data analysis at VHB

Crash analysis and interactive dashboards for 15+ Safety Action Plans across North Carolina, Colorado, and New York.

2023–2024 Data analyst

At VHB I was the data person on Safe Streets for All, the federal program that funds cities and states to figure out where people are getting hurt and actually do something about it. Over a year and change I worked on more than fifteen of these Safety Action Plans across North Carolina, Colorado, and New York.

The raw material was always the same kind of mess: years of statewide crash records, in a slightly different format every time, that I had to clean, normalize, and stitch together with roadway and equity data. The fun part was turning that into something a planner or a city council member could actually sit with. I built interactive Power BI dashboards where you could see where the severe crashes cluster, which ones overlap with historically underserved neighborhoods, and whether a proposed fix would land where the risk really is.

After the third or fourth plan I got tired of rebuilding the same thing, so I made reusable data models that let us spin up a new jurisdiction’s analysis without starting from scratch. That’s the instinct I keep coming back to: if I’m doing it by hand twice, I’d rather build the tool.

resume

open as PDF ↗