I'm Long.
I keep up with
AI so you don't
Solo founder of RDTS. Nearly a decade shipping production software — started in 2018, before Midjourney existed and well before LLMs could write usable code. I build with AI daily: agents, Claude/GPT, MCP servers, custom harnesses. That's the job. Funded startups and growing companies hire me as their fractional AI lead so they don't have to keep up themselves.
I started writing software for a living in 2018 — before Midjourney existed, before ChatGPT existed, before LLMs could write usable code. I came up traditional: real backends, real frontends, real databases, no AI lifting the keyboard for me. By the time the first useful LLMs landed in 2022 I had four years of production-grade code under my belt and a very clear sense of what AI was good for and what it wasn't.
I spent those early years as the in-house web developer at two marketing agencies — first as a junior, then as the engineer the founder went to when something had to actually work. Both shops sold creative work as the headline product. Both quietly depended on whatever automation glue I could rig up to make the operational side stay standing.
What I kept watching, at both shops and at every team I've worked with since, is the same scene: the firehose never stops. New models, new harnesses, new integrations, new capabilities announced every two weeks. Most operators don't have time to evaluate any of it. They're running the company. So the team misses what matters, adopts what's hyped, and ends up with a pile of seat licenses that don't do much.
That's the problem a fractional AI lead solves. I track the firehose for you. I know which releases matter for your stack and which are noise. I audit what your team is using, build what should exist, wire in the integrations that compound over time, and show up on retainer so there's always someone who keeps up — without you having to hire a full-time head of AI.
That's RDTS. The offer is a fractional AI lead on monthly retainer for funded startups and growing companies. Harness-agnostic, model-agnostic, opinionated about matching the right pieces to the use case.
Talk to the duck if you want to hear what that looks like for your specific situation.
Keeping up is the job.
The AI release calendar doesn't pause for your roadmap. Part of what you're paying for on retainer is someone who reads the announcements, filters the noise, and tells you what actually changes your stack.
Harness-agnostic, on purpose.
Claude Code, Cursor, Codex, OpenClaw, Cline, Aider — they're all good for something. I pick the one that fits the team and the work, not the one I have a discount code for.
Source code transfers to you.
Skills, agents, MCP servers, software builds — all of it ships into your repo, your infra, your accounts. Take it anywhere later.
Solo, on purpose.
No agency layer between you and the engineer. The person you talk to first is the person doing the work. That's the entire point.
Specific over generic.
I tell you which harness, which model, which integration — and why. You hire me for opinions, not vibes.
Honest no-list.
Foundation-model pre-training and cleared work — I refer those out. Everything else in the fractional AI lead lane, I'll quote honestly.
Ships things.
Keeps up with AI.
The two things you want from a fractional AI lead: real engineering chops and genuine fluency with current AI. Three builds that show both:
Lee De Card
lead dev · 2024AI-driven creator booking platform. Calendar integration, automated outreach, intake workflows. Built on a small team where I owned the engineering side end-to-end.
Dragon Wagons
site + tooling · 2024Marketing site + content automation backend. The kind of "we need a real engineer for the ops layer behind our public-facing site" project that's a textbook RDTS engagement.
Easiest way
is to talk.
Describe your setup, your team, your stuck point. The duck reads the shape, names a price in plain English, and sends a scoping brief to you and to me. Five minutes, no funnel.