Questions, answered.
Everything about security, safety, and how it runs day to day.
The assistant is inbound-only — it never sends bulk cold DMs or starts conversations, it only replies to people who message you first. That keeps your account far safer than spam-style automation. No automation is completely risk-free, so we also pace replies naturally and give you full control.
No. It's locked to the prices and policies in your settings. If it doesn't know something, it tells the customer it will check rather than inventing an answer.
Type your take-over command (e.g. //stop) in a specific chat and the assistant goes silent for that customer while you reply. Type your resume command (e.g. //start) to switch it back on. It keeps reading and remembering the whole time.
Your session and API keys are encrypted at rest (AES-256 / Fernet). The key lives only in our environment, never in the database, and tenant data is isolated with row-level security. You can disconnect your account in one click.
It's warm and fast, and honest when asked — it helps as your assistant rather than pretending to be a human. That keeps you compliant and your customers' trust intact.