sudo make homelab
What is actually running in my homelab — the hardware, the network, the platform, and the honest gap between what is here and what is planned.
Real builds. Real failures. Real lessons.
Where AI-generated ideas become production systems.
8 posts
What is actually running in my homelab — the hardware, the network, the platform, and the honest gap between what is here and what is planned.
The build log for sudomakevibe.com — stack decisions, six things that broke, and what a corrupted disk taught me about deployment pipelines.
R7 — the Carrier-Grade AI Stack. Seven layers for regulated AI deployment. Indemnity covers provenance. Insurance covers performance. Regulated AI needs both.
Part two of the AI cost series. Technical debt, false confidence, and the quiet atrophy of engineering skill. The costs that determine whether AI pays off or breaks your systems.
Everyone talks about how fast AI lets you build. Few talk about what it actually costs. Part one — the costs that show up on your invoice.
High-fidelity work requires high-fidelity documentation. Here is why this site exists as a blog and not a two-hour podcast.
Correctness, reliability, security, maintainability — these are not features. They are disciplines. AI doesn't ship them. You do.
This is not another AI hype blog. It is a working lab — documenting what actually happens when AI-generated ideas meet production systems.