Learn from me
Read the Field Notes
Twenty short posts from running an engineering org where AI writes most of the code - review, QA, agents, adoption, hiring - and a six-part series on the AI Platform Map.
Context Is the Architecture Now
RJ Lindelof - building AI-native systems where versioned Markdown drives the SDLC
This is the builder's side of the work. I run a multi-model agent stack against real production codebases, and the thing that decides whether it ships is never the model - it is the context you hand it. Specs, architecture decisions, and intent live in version-controlled .md files that humans review and agents consume. The Field Notes and the approach behind them are what that looks like in practice: the pipeline stages, the guardrails, the failure modes I hit, and the numbers that came out the other side. If you want the org-leadership framing of the same work, that is rjl.ai.
- 20+Years engineering leadership
- 99.95%SLA delivered
- 70%Faster onboarding
- 5xDeploy frequency
- 65%Tech debt retired
What This Site Is
Same practice, different altitude.
RJLAI.dev is where the implementation details live. Agentic delivery is not a model choice, it is a systems-design problem: what the agent can see, what it is allowed to change, what proves the change is correct, and what happens on the run where it is confidently wrong. Those four questions decide whether AI is a delivery system or an expensive autocomplete.
The pieces I keep coming back to: context as a versioned artifact rather than a prompt, intent specs that say what to value and not just what is true, review as the deliberate constraint once implementation stops being the bottleneck, and evals wired in before anything is allowed near production. The field notes go through each of them with the arguments I lost before I won them.
What Brings You Here?
I know specific things about putting AI to work in software development, because I do it every day. Pick the door that fits.
Hire me
Work With Me
Engineering leadership that moves AI from pilot to production in regulated B2B SaaS - the delivery system, the organization around it, and the platform underneath - with numbers to back it.
Connect and talk
Compare Notes
Working through the same problems - AI-native SDLC, agentic delivery, platform modernization? I enjoy comparing notes with other leaders, and the door is open to the right role or project.
From the Field Notes
Real-world writing on AI in software development: what changed when AI started writing most of the code, and what it took to trust the result.
The newest series, in six parts The AI Platform Map AI made generating software cheap; the series is about making it cheap to believe the result. Six posts walk the nine layers of an AI engineering platform: written intent and a durable workflow that outlives the agent, ephemeral agents with their own scoped identity, a model you can swap and measure, three verification layers that send a person only what needs a person, and telemetry that counts trustworthy work instead of AI usage. Start with part 1
When Most of the Code Is AI-Written
Current practice, 2026 A high share of AI-written code measures keystrokes, not engineering. The scarce skill is deciding which code should exist at all. Read the post
An Operating Model, Not a Tooling Purchase
Operating guide, September 2026 Authoring got cheap and verification is the constraint. Ten rules, the practices behind them, and how the rollout fails. Read the post
What the Data Says in August 2026
Industry data, published 2025-2026 Adoption is near-universal and trust is falling. DORA, Faros, METR and GitClear on where AI-assisted delivery actually stands. Read the post
The Ideas Behind the Work
Three shifts decide whether AI ships or stalls.
The Context Revolution AI operates on context, not abstraction. Versioned Markdown is the substrate, and how well you feed context to the models is the advantage - not the model. Read the idea The Intent Imperative Context tells an agent what is true; intent tells it what to value. An agent ships with a written intent spec, enforced by evals, or it does not ship. Read the idea The AI Bottleneck Shift Speed up coding and the constraint moves to requirements, review and QA. The win is the whole value stream, not the fastest phase. Read the idea
Proof, Not Promises
Four engineering organizations, one pattern: pick the right scope, prove value in shipping increments, and instrument everything.
"A standout technology leader who balances high-level strategy with a true roll-up-your-sleeves work ethic. RJ was a driving force behind our most critical technical transitions - most notably the shift toward an AI-native SDLC."
Julie Gniffke, SVP of Engineering, hc1 - read it on LinkedIn
Let's Compare Notes
Working on AI-native delivery, hiring for it, or looking for someone to speak about it? I read every message.
