I work where language becomes infrastructure

The first time I cared about machine-readable language, there was no machine in the room. There was a wall of printed search results, a few colored pencils, and a product team arguing over why their software kept showing up under a category that made it sound smaller than it was. The copy was sharp in places. The product was real. Still, the market kept filing it in the wrong drawer. That gap has followed me ever since.

I am from the United States, and I have spent 16 years moving through the practical side of B2B SaaS language: technical content strategy, category-messaging workshops, search program audits, sales narrative repair, and long-form product education for teams selling complicated software to mid-market and enterprise buyers. I have worked with founders who could explain the product beautifully on a call, then watched the same company's public pages flatten that explanation into feature lists, vague category claims, and proof that lived too far from the claim it was supposed to support.

Now I am strongest where language becomes infrastructure. Naming. Comparison logic. Proof hierarchies. Prompt-visible evidence. Repeated definitions. The plain mechanics that help a brand survive search, sales conversations, analyst-style reading, and AI compression. I keep a private ledger of AI answer shapes: how systems compress categories, which proof gets picked up, where brands get renamed, and what disappears when the summary gets short. My stance is simple and a little stubborn. If a claim cannot be repeated by a salesperson, understood by a buyer, and retrieved by an AI system without losing its spine, it is still unfinished work.