Per-organization isolation
Another customer's conversations and docs are not in your workspace — and yours are not in theirs.
How isolation works
Each organization gets its own data space for knowledge, conversations, and records.
Only the people and agents you allow can read or write inside that boundary.
Subagents only get the knowledge, tables, and app connections you attach.
Website widgets authenticate access so a pasted snippet cannot become a free-for-all API.
Capabilities
Practical boundaries buyers and IT teams actually ask about.
Another customer's conversations and docs are not in your workspace — and yours are not in theirs.
Uploaded policies and product docs are searched inside your org boundary.
Leads, tickets, and feedback rows live with the organization that collected them.
A product-lookup helper does not need email send access — and should not get it.
The public widget is built so visitors talk to your agent without exposing admin power.
Team members see what their role allows — not every secret key and every table by default.
In practice
Short answers without vendor jargon.
Embage is built around organization isolation so your operational data is not casually mixed with other customers.
Visitors talk through the agent under the access rules you set — they do not get a raw dump of your entire library.
Builders and admins in your organization. The public embed cannot reconfigure your agent.
App connections are authorized per organization and attached only to the helpers you choose.
FAQ
Embage is designed so each organization's conversations, knowledge, and datastore records stay isolated from other organizations.
No. The embed is meant for visitor conversations with your agent under controlled access — not unrestricted access to your admin data.
Yes. Knowledge sources, datastore tables, and connected apps are attached intentionally — especially via focused subagents.