Subscribers hand Tars their own strategy, brand, operations and reference material — so it answers as an advisor who already knows the business, not a general chatbot. It reads that context on every message, and it never leaves your workspace.
On the public Tars you're talking to the general model. Company context is part of Tars Enterprise — the plan built for a business that wants an advisor grounded in its own material.
On the public Tars, memory lives with your session: the recent back-and-forth of this conversation, cleared after 30 minutes of quiet. Nothing about you is kept after that.
Standing memory that follows you between conversations comes with any Tars plan — Individual, Team, or Enterprise.
When on, your question — and only your question — is sent to a search service. Nothing else about this conversation leaves.
Tars is a product of Creative by Request. For customer service, general questions, or enterprise enquiries, the one contact channel is hello@elevatoretiquette.com. There is no phone line.
This demo runs on Creative by Request's own hardware; your messages are not sent to any third-party AI service. Conversations here are kept in memory only and are cleared after 30 minutes of inactivity. The one exception is web search: when you switch it on in System, the text of that question — and nothing else — is sent to a search provider.
Tars can be wrong. It is a thinking partner, not a substitute for professional, legal, or financial advice — check anything you intend to act on.
© Creative by Request. All rights reserved.
Private, locally run, and never metered by the token.