Industry workflows, rebuilt with AI agents.

I take one industry's real workflow at a time, rebuild it with AI agents, and show the whole thing.

AI can already do more than most professions have figured out how to use. That gap — between what the models can do and what actually runs in your day-to-day work — is what this is about.

Getting agents to finish real work under real constraints is where most AI projects die. That's the part I work on. English is my second language, so I keep the prose plain.

Working pieces from rebuilding real industry workflows with AI agents. Prompts, skeletons, and the judgment calls behind them.

Everything demonstrated in a video is here. If you saw it, you can run it.

From generic AI to industry workflows that ship.

• One industry at a time. Take a real workflow, rebuild it end to end, show the setup, the failures, the numbers, and the parts that still need a human.
• Only what I've run. If I haven't broken it, I won't write about it.
• Every model release is another round. Same workflow, new generation, what changed and what didn't. Old posts become the baseline rather than going stale.
• No vendor money. No sponsorships, no affiliate links, no paid reviews. Nobody pays me to say a tool is good.

• 🌐 aiworkflowpro.com — write-ups, with the numbers and the failures
• 𝕏 @aiworkflowprolk — what I'm rebuilding right now
• ▶️ YouTube — one workflow per video, start to finish

Public repos here are clean snapshots — tools, examples, and docs meant to be searched, reused, and inspected safely. Product source code, work-in-progress experiments, and internal automation stay private by default.

_{The models get better. Someone still has to use them.}

— Leo