Kopai: what owners actually say

Users explicitly worry that when an agent gives a wrong or outdated answer under an expert's name, it's unclear who is on the hook — the expert or Kopai. One commenter notes per-hour consulting has a built-in correction loop that per-message lacks.

A user points out that the eval gate fires on edit, but the failure mode is that experts upload once and step out of the loop entirely. Eighteen months on, the expertise may be stale with no mechanism to catch it.

A commenter asks how quickly the orientation feature can bring an agent back on track if it starts drifting from the original goal midway through a long workflow — suggesting this isn't well documented.

Users want to know what happens to conversation state between messages — whether an agent retains context from several messages prior when a follow-up is asked later.

Multiple users call the shift from per-hour to per-message a smart unlock, with one saying it makes more sense than subscriptions for this kind of tool.

Co-founder highlights a custom Gen-UI library used in chat and agent answer generation, presented as a standout feature.

Users acknowledge the core problem Kopai targets is genuine — experts can only sell knowledge one hour at a time and hit a hard ceiling when their calendar fills up.