3. Your content and connected workflows
Customer Content includes documents, memories, extracted facts, relationships, questions, saved conversations, chat messages and content imported from authorised connectors. Browser features may submit the context you choose to use. Recording and transcription features process audio and the resulting transcript when you activate them.
Enkeon-E can process shared agent progress, questions, decisions, feedback and proposed procedures where enabled. Review generated answers and procedures before relying on them. You are responsible for notices, permissions and recording consent required for the people whose data you submit, including meeting participants.
Project graphs use per-project engine storage. Shared control-plane storage also holds account records, connector records and chat content. Access depends on project and team permissions; sharing a public agent or embedding it in an application can make the content exposed through that agent available to its users. Check permissions before publishing.
Connecting AI tools.When you connect an AI tool such as Claude, ChatGPT or Cursor to a project — by signing in, or with a project key — the tool can look up and save that project's context. Enkeon receives only what each tool call sends: the words looked up and the text saved. It never receives the conversation itself. Each call is recorded with the tool used, the time, the name of the AI tool and the branch, so you can see in the Console what was read and saved. Saved items are kept like any other Customer Content (see Retention and deletion). You can disconnect a tool from its own settings at any time.
Your content is processed to provide the features you use. It is not a shared public knowledge pool. Project sharing and public-agent access follow the settings you authorise.