- The problem
- Dispatch tools show you the load, not the money, and the numbers live in five places that never match.
- What I built
- One system for quoting, dispatch, billing, driver pay, a live fleet map and an audit log. Every offer is priced against that truck’s own cost. A rate confirmation is read into a draft load by an AI service, and a person checks every field before it is saved. Spend limits cap what the AI reading can cost.
- Built with
- Next.js, TypeScript, Prisma and Postgres with row-level security, with automated tests.
- How I checked it
- Run against invented data at 1,500 tractors, 5,000 trailers and 50,000 loads, with most pages opening in about a second.
I build the tools behind the process.
Nearly five years in freight operations, and I build the software that takes the manual work out of it: a dispatch desk, a driver app and a validated booking platform. Each one is a solo build, written up as a specification first and checked with automated tests.
The TrueLane demos run on a sample fleet: every name and figure on screen is invented.
Selected work
- The problem
- Owner-operators say yes to loads that look good and lose money, because a rate means nothing without the miles, the costs and the empty return.
- What I built
- An offline-first app that sets every offer against the driver’s own cost per mile and states the basis it used, reads rate confirmations into loads, and carries the work through trip tracking, invoicing, and tax and expense records.
- Built with
- Expo and React Native, TypeScript, Postgres with row-level security, thousands of automated tests.
- Where it stands
- In TestFlight beta, not in the App Store yet.
Booking-plan platform
Internal tool, Ashley Furniture Distribution Services- The problem
- Backhaul booking ran across five disconnected tools, with duplicate entry and recurring data errors.
- What I built
- I mapped the workflow with DMAIC, baselined the admin time and the errors, then designed and built a replacement: a validated schema, document ingestion with human review, per-load documentation gates, cancellation reason codes and full audit logging, backed by an ~11,500-word specification.
- How I checked it
- More than 500 automated regression tests. In the same role I rebuilt the regional KPI workbook, found a silent formula defect that left two lanes out of the regional totals, and added automated reconciliation checks.
- Result
- Duplicate manual entry across systems was eliminated. The region’s revenue per mile went from $0.04 to $0.72 against a $0.65 goal. This is an employer’s internal tool, so there is no public demo.
How I work
Specification first
I write down what the system must do before I build it, so the build has something to be checked against.
Tested, not trusted
Every calculation has automated tests and every table that feeds a number has a reconciliation check.
A person signs off
Where AI reads a document, a person reviews every field before it is saved.
Built solo, with AI tools
I build alone using AI-assisted tooling, from written specifications. I review and test what it produces.
Open to
Full-time roles in logistics systems, automation and analysis, in the DFW area or remote, and fixed-price builds for small operators. The quickest way to start is fifteen minutes on a call.