David L. Bowman
I lead engineering teams. I grew the engineering team at Bright Development to 72 engineers and the business to $8M in annual revenue. Today I lead the team building OpenRouter's platform for AI agents.
Most of my team’s code is written with AI agents. The engineers own the design, review every change and write the tests.
Austin, TX
What I've built
72 engineers at Bright Development
I co-founded Bright Development and grew its engineering team from the first hire to 72 engineers in the US and abroad. I ran hiring, team structure and delivery, and we shipped more than 250 products in three years.
$0 to $8M in annual revenue
We grew Bright to $8M in annual revenue in three years and were profitable every year. I owned headcount, the engineering budget and what we committed to deliver.
A team that builds with AI agents
I lead the Ori team at OpenRouter. We build deployable agents, OpenRouter inference for Claude Code, Codex and other coding agents, and the vault, managed machines and evals around them, using the same agents to do it.
Software that's still running
As CTO of Ocean Block Capital, I built the system the firm used to source properties and decide what to bid. More than 60 brokers used it, it handled more than $250M in acquisitions, and it still runs today, unchanged. I also managed the firm's 12 acquisitions associates.
How I lead
- 01
Agents write the code, engineers own it
On my teams agents write most of the code, often several at once on separate branches. Engineers write the spec, decide which model to trust with which job, review every change and own the tests.
- 02
Stay in the code
I still write code most days, mostly TypeScript with Effect. Decisions about scope and estimates are better when the person making them knows the codebase.
- 03
Own the numbers
At Bright Development I owned headcount, the engineering budget and what we committed to deliver. The business was profitable every year.
- 04
Build things that outlast you
The system I built at Ocean Block Capital handled more than $250M in acquisitions and still runs today, unchanged.
Start here
Agent Harness 101
Every coding agent is a short loop: ask the llm, run the tools it asks for, and stop when it asks for nothing. The harness is everything around that loop that keeps it safe and durable, and each of its primitives (session, thread, turn, tool, permission and agent) is written here as a real schema. Ends with a scripted demo that follows one message from the moment you press enter to the final answer.
Take the Keys Away From Your Agent
A process that runs model-authored code should hold no credential worth stealing and have no route to anywhere it could send one. Builds that from the bottom up: what TCP, HTTP and TLS actually are, how certificate authorities decide who you believe, and what a proxy can and cannot see. Ends with an interactive walkthrough of one request through every process that touches it, which you can rerun on a flat network to watch the whole thing become irrelevant.
AI Workflows That Don't Rot Your Codebase
AI is leverage only inside a disciplined loop — plan, devil's-advocate, execute, post-mortem, improve your skills. Let it push ahead unsupervised and the same tool quietly wrecks the project: drifting to mimic old code, writing docs that pass typecheck but lie, leaving half-refactors and parallel types behind. A field guide to the loop that ships and the anti-patterns that sink, with downloadable skills and a real brand-book deliverable.
Get in touch
Have a question about something I wrote, or want to talk? Send me a message.