Hermae

In 2024 we blended LLM generation with living design systems. Public demos drew interest from design-system maintainers at Fortune 500 companies who needed help getting new teams onto tokens, components, and documentation without a long training cycle.

Hermae retrieved against system source and generated an assistant on top. We shipped an API and a dashboard so teams could connect their own assistants, and we scaled the build to about ten contractors when the product needed more hands.

The product held up, but the sale did not. Enterprise security and compliance reviews slowed every conversation, and for most buyers the tool stayed useful but non-essential once procurement got involved. We wrapped the company in February 2025.

What remains useful is the RAG and vector-store patterns, along with a clearer read on when a well-built AI product still cannot clear an enterprise buying motion.