Case file 04 · AI-accelerated engineering
The only case file here
you can audit yourself.
The other three are my word, corroborated by people who were in the room. This one is a commit history. If you want to know whether the pipeline I would build your product on actually works, the honest answer is to go and look at what it has already built.
Subject
My own portfolio. Twenty-eight repositories, built solo on an AI-accelerated development pipeline, alongside a full advisory practice.
The situation
Thirty-eight years of security work gives you strong opinions about how software should be built and no time whatsoever to build it. That was true for most of my career. The tooling changed that, and I wanted to know by how much — not as an opinion for a conference talk, but as a measured thing. So I ran the experiment on myself: how much can one operator actually ship, working solo, without a floor of engineers underneath?
What is actually here
Twenty-eight repositories and more than five thousand six hundred commits, the earliest dated June 2025. Compliance platforms, an attack-surface scanner, athletics logistics, an endpoint-monitoring tool for consumers, several games. Five are live on their own hostnames right now, serving a real product rather than a holding page:
ecombine.appCombine and athletics dataetrax.appEquipment tracking for athletic programsrecon.threattape.comExternal attack-surface reconnaissancerisktape.ioCyber governance and compliance evidenceapp.sourceiq.infoContent provenance and threat intelligence
What is not here
Twenty-eight shipped products. That is not the claim and it would not survive five minutes of scrutiny. Most of those repositories are in active development, several are explicitly not ready, and at least one does not currently build. Every one of them is under the same standard I would hold your product to, which is exactly why I can tell you which is which. A portfolio that claims everything is finished is telling you something about the person, not the work.
Why it is relevant to you
Because the pipeline is the product I would be selling you, and this is it running unsupervised for fourteen months on my own money and my own reputation. Not a demo built for a pitch, and not a case study about somebody else’s team. The same approach that produced this is the one that produces your MVP — with the difference that yours also has to pass an enterprise security review, which is the part most AI-accelerated builds discover too late.
Result
What this means for you
Ask any vendor pitching you AI-accelerated delivery to show you what they have built with it for themselves. Most cannot, because the pipeline is something they sell rather than something they use. It is a fair question, it takes one link to answer, and the answer tells you more than any proposal will.