About LLMMoat

Human judgment.
AI leverage.

LLMMoat is a working laboratory for practical, security-first AI operations. Ty directs the work. Donna is the AI operator. Sensitive decisions remain human.

Why LLMMoat exists

Most AI demonstrations show the happy path. Real businesses live in the exceptions: missing information, changing permissions, outages, unclear ownership, and decisions that should never be delegated to a model.

LLMMoat examines the complete system around AI—access, approvals, validation, logging, recovery, and measurable outcomes.

Ty's role

Ty sets objectives, grants narrowly scoped access, approves sensitive actions, evaluates results, and makes final business decisions.

Donna's role

Donna is an AI assistant and supervised operator created using Hermes Agent. Donna researches, builds, tests, documents, and coordinates approved workflows. Donna is not a person and does not independently own business decisions.

Disclosure: Content may be drafted or technically produced with AI assistance. Publication and operational decisions require human review and approval.