Work
What I've built
Each project in four lines: the problem, my role, the approach and the outcome. Numbers come from each project's own page.
filter by what you care about ↓
-
Talk · AI Engineer NYC 2026
- Problem
- Agents made coding fast, but decisions, sign-offs and context around the code stayed slow.
- My role
- Designed and ran the war room; gave the talk; wrote the GEICO Tech Blog post.
- Approach
- One room of decision makers and a driver, with an agent fleet (FleetKit) planning work as issues people approve.
- Outcome
- An agent-callable API running in UAT by the end-of-day demo, on a $1,000 token budget.
Speaking
-
Explainer
- Problem
- Data center debates in Virginia mix power, water and land figures without a shared baseline.
- My role
- Researched the sources and built the explainer and scenario model.
- Approach
- A flow diagram of electricity, water and air, cited state and county figures, and a four-path scenario calculator.
- Outcome
- 78 sourced figures and a model calibrated against JLARC's measured 2023 water use.
Civic Tech Reports
-
Report · PlanRVA
- Problem
- The region's new long-range plan is spread across PDFs and a map service.
- My role
- Parsed the plan's project tables and built the report.
- Approach
- A money-flow diagram, a timeline by time band and a filterable map of every project.
- Outcome
- 121 new projects and $7.9 billion traced from revenue to mode, reconciled to the report's totals.
Reports Civic Tech
-
Hack for RVA · 2026
- Problem
- Hackathon ideas scatter across teams, repos and demo slides after the weekend.
- My role
- Chief Hacker and co-founder of Hack for RVA.
- Approach
- One index of every idea, labeled by pillar, by where it came from, and by AI-persona review verdicts.
- Outcome
- 75 ideas in one place, 39 built during the weekend and 31 marked ship.
Civic Tech
-
Hack for RVA · Research
- Problem
- Teams needed context on each city priority before writing code.
- My role
- Chief Hacker and co-founder of Hack for RVA.
- Approach
- A research repo per Mayor's Action Plan pillar, each answering the same nine questions.
- Outcome
- 396 research files across 7 pillars and 14 problem statements from city experts.
Civic Tech
-
Hack for RVA · Open toolkit
- Problem
- Other cities want to run a civic hackathon without starting from zero.
- My role
- Chief Hacker and co-founder of Hack for RVA.
- Approach
- A guide plus planning skills that turn blank templates into a city's own plan, scheduled by dependency.
- Outcome
- A 42-page guide, 19 planning skills and 79 templates.
Civic Tech
-
Analysis · Workforce
- Problem
- It is unclear which Richmond jobs AI exposure actually threatens, and where those workers could go.
- My role
- Personal project; built the analysis and the site.
- Approach
- Matched occupations to AI exposure scores and tracked BLS employment from 2023 to 2025.
- Outcome
- Exposure did not predict decline; 28 pathways to better-paid, growing jobs.
Civic Tech Reports
-
Flying Pig Labs · Tool
- Problem
- Turning research into a fundable startup idea takes weeks of manual work.
- My role
- Forward deployed engineer and co-founder of Flying Pig Labs.
- Approach
- A pipeline that generates ideas from one prompt, has AI investors review the best, and opens a repo for the winner.
- Outcome
- 60 ideas, 54 investor reviews and an approved project repo in 8.5 hours.
Prototypes
-
Flying Pig Labs · Platform
- Problem
- Sprint teams wait days for a cloud environment and rarely clean it up.
- My role
- Forward deployed engineer and co-founder of Flying Pig Labs.
- Approach
- A live environment per team on day one, AI inspection of the prototype, and teardown when the sprint ends.
- Outcome
- 6 AI inspection specialists per environment, 12 architecture decision records, a $150 monthly target per install.
Prototypes
-
Flying Pig Labs · Workshop
- Problem
- People with a validated idea stall before anything is live.
- My role
- Forward deployed engineer and co-founder of Flying Pig Labs.
- Approach
- Nine guided steps from idea to a deployed landing page that collects sign-ups.
- Outcome
- About two hours, start to finish, with 8 resource guides.
Prototypes
-
Prototype · Lending ops
- Problem
- A total-loss car claim touches many lender systems, and money must never move without a person.
- My role
- Personal project; built it.
- Approach
- One queue across six mocked systems, with AI that reads and coordinates but never approves.
- Outcome
- 6 human approval gates and a payoff that reconciles to $0.00.
Prototypes
-
Course · Exam prep
- Problem
- Exam prep for Claude Code is reading-heavy and light on practice.
- My role
- Personal project; built it.
- Approach
- Read the concept in the app, then do the exercise in your own editor with Claude Code.
- Outcome
- 21 days, 42 exercise files and 5 exam domains.
Prototypes
-
2010 Appalachian Trail
- Problem
- My 2010 thru-hike journal lived only on a website.
- My role
- Personal project; built the extractor.
- Approach
- Pull every entry into one file, keeping date, start, destination and miles.
- Outcome
- A source for a printed book.
Personal
-
Music
- Problem
- A community needed to search and share sacred chants at the exact moment that matters.
- My role
- Personal project; built it.
- Approach
- A serverless site, API, data model and frontend.
- Outcome
- 169 commits across 4 repos in 3 weeks.
Personal
-
Decisions
- Problem
- The preferences people state often differ from how they actually react.
- My role
- Personal project; built it.
- Approach
- Five guided phases, each a service, that test stated preferences against reactions.
- Outcome
- 5 phases, from the repo; the live app is offline today.
Personal
-
Writing
- Problem
- Lessons from the work stay inside the teams that learned them.
- My role
- Author.
- Approach
- Short essays, first published on LinkedIn, now collected with an RSS feed.
- Outcome
- 13 essays from 2021 to 2026.
Writing
What people say
A short quote from someone Ford has worked with goes here.
Name · Role, OrganizationA short quote from someone Ford has worked with goes here.
Name · Role, OrganizationA short quote from someone Ford has worked with goes here.
Name · Role, Organization