An engineer musing in the clouds

Do you think days are shorter post-COVID 🦠?

I'm trying to make the most of it by writing about my thoughts and strong emotions and such.

Featured

• Genie Ontology: The Real Product Is Lock-In

  Published:Jun 18, 2026 |  at 04:00 PM

  Genie Ontology, Horizon Context, the IQ family — six context layers shipped in four days. The agents are disposable and the models are commoditized. The semantic graph underneath is the part that doesn't get swapped, and that's exactly why every data platform is racing to own yours.

• Context Is Necessary. It Isn't Enough.

  Published:Jun 17, 2026 |  at 03:00 PM

  The whole data industry is racing to ship context graphs — Genie Ontology, Horizon, the IQ family. Context is real and necessary. But it tells an agent what's true, not whether it did the right thing. The open problem is verifiable execution.

• Production AI Is Mostly Boundary Design

  Published:Jun 16, 2026 |  at 03:00 PM

  Production AI works when the model owns the ambiguous middle, not the whole workflow. The lesson shows up in a lead dedup case study, but the real story is boundary design.

• Agents Shouldn't Grade Their Own Homework

  Published:Jun 9, 2026 |  at 03:00 PM

  GitHub built an impressive stack to validate non-deterministic agents — a prefix-tree graph, dominator analysis, a multimodal LLM. I needed the same thing for Grove's agents and found the hard part didn't exist. The reason is the whole argument for structured trajectories.

Recent Posts

• Security & Compliance: AI Workflows for Regulated Industries

  Published:Jun 12, 2026 |  at 03:00 PM

  Grove is a DAG-based workflow engine built for regulated industries. One-shot execution, per-node audit trail, AES-256-GCM encryption, air-gap deployment, and provable data disposal for organizations with strict data custody requirements.

• The Forward Deployed Engineer Boom Is an Admission

  Published:Jun 5, 2026 |  at 03:00 PM

  a16z calls it the hottest job in tech and the AI labs are hiring them by the hundred. The forward deployed engineer isn't a new kind of engineer — it's an admission about where the hard part of AI actually lives.

• The 20% Who Actually Make Money on AI

  Published:Jun 2, 2026 |  at 03:00 PM

  Real AI profit exists — it's just concentrated in a fifth of companies that all look strangely alike. The winners' profile turns out to be almost the inverse of how everyone else is adopting the technology.

• The Speedup Trap: What AI Did to Engineering Expectations

  Published:Jun 1, 2026 |  at 04:00 PM

  AI made writing code faster. Organizations turned that into a mandate. The problem is that the speedup is the one part of the job that got easier — and it's the part nobody was struggling with.