RJ Lindelof
Outtakes: The Mistakes and Lessons Behind the Resume
4 min read
Why a CEO Site Has an Outtakes Page
A resume is a highlight reel. Every other page on this site tells you what went right, with percentages. This one covers the rest.
Here are the lessons I learned the slow way, the opinions I had to trade in, and a few facts that keep me humble. Nothing here is at anyone else's expense. I had plenty of material without borrowing any.
Lessons learned
Mistakes and Lessons Learned the Slow Way
It took me thirteen years to stop writing code for a living
Most executive bios skip quickly past the individual contributor years. Mine lasted thirteen. I now describe this as "deep technical grounding," which is what you call a slow start once it has paid off. The lesson is real, though: I can still tell when an estimate is fiction.
Making coding faster did not make delivery faster
AI sped up the typing. Then review, QA, and backlog grooming choked, exactly as a 1984 book about factory bottlenecks said they would. What I do now: plan for the constraint to move before the tools go in, and make review slower and more expensive on purpose.
The expensive model was doing the cheap job
I profiled an AI workload and found it was using under 2% of the frontier capability it was paying for. Moving it cut inference cost 90%. Flattering as a result, less flattering as a reminder of how long a bill can go unquestioned. What I do now: profile first, and assume the right model is the cheapest one that clears the bar.
Buying licenses is not adoption
AI adoption in engineering is not a tooling rollout. It is a behavior and trust problem, and the skeptics on the team are usually the ones worth listening to. What I do now: start where the pain already has a budget, and never report adoption percentage as success.
What I got wrong
Opinions I Have Traded In
- Prompt engineering. It turned out to be the old conversation. It was never really about the words. It was about intent, and intent is harder to write down than a clever prompt.
- AI and senior engineers. AI did not shrink the need for senior engineering judgment. It made judgment the whole job. Four weeks after I wrote that, I was writing a follow-up, which is roughly the shelf life of any confident statement about AI.
- Remote work. Remote was never the variable. Clarity was. Distance does not kill teams. Ambiguity does, and it is much harder to blame on a time zone.
Humility, enforced
Arguments I Lose, and the Meetings I Deserve
The teams I have built will challenge me, and I will lose those arguments openly when they are right. This is by design. It is also occasionally inconvenient.
I have a name for a kind of meeting I have attended and, I assume, caused: RANCI. It is RACI with an N, which stands for "Nobody knows why they're here."
And one of my own lines came back to me in a LinkedIn recommendation. An engineering manager wrote that I prodded him past his caution by asking, "If there's so little risk, why wait?" It is a good question. It is also an annoying one to hear from your boss.
Anti-resume
A CEO Since 1994, in Numbers Nobody Asked For
0
Graduate degrees. I have not found a role in 20+ years where this mattered.
13
Years as an individual contributor before I led a team.
1
Publication with jQuery in the title.
400+
Pages in my guide to AI-assisted development. Brevity is on the roadmap.
Also on the record: I keep a deliberately terrible 1990s-style homepage, complete with "Best viewed in Netscape Navigator 4.0+" and "Created with Notepad & Love." It lists my age as "Old enough." You can see it at rjlindelof.com, RJ's retro personal homepage (external).
AI disclosure
How This Site Was Built
It would be strange to preach AI-native delivery and then pretend I hand-typed every page. I did not. This site was planned, written, and built with AI agents working from my own source material: my resume, my writing, my LinkedIn recommendations, and the previous version of this site.
The rules were the ones I give any team. The facts are mine and come from my records. The testimonials are quoted word for word. The agents were told to invent nothing, and a human owns the merge. If you find a sentence that sounds like autocomplete with a hoodie, that one got past me, and I would like to hear about it.
The serious bit
Leadership Lessons Learned: Leaders Go First
Admitting a mistake early is cheaper than defending it late, and teams only do it when the leader goes first. Psychological safety is not a soft skill. It is what makes teams call out the questions that prevent the next outage. That is the whole reason this page exists.
The principles behind it are on the leadership philosophy page, and the flattering version of the story is on the about page.
Spotted a mistake I missed?
I collect them. Send it over and it may end up on this page.