We built an AI that asks before it acts. Here is the architecture.

Description and use

Weio’s architecture delivers trustworthy AI by grounding answers in source citations, requiring explicit approval for actions, and maintaining an immutable audit trail, reducing accidental data changes.

Problem

Users must manually copy AI answers between tools, leading to slow coordination and potential errors when AI interacts with live systems.

Outcome

The system delivers trustworthy AI interactions by grounding answers in source citations, requiring explicit user approval for actions, and maintaining a comprehensive audit trail, reducing risk of accidental data changes.

Workflow

• 1. User asks a question in the AI tool.
• 2. AI retrieves relevant sources from connected apps and generates an answer with attached citations.
• 3. If the user wants an action, the AI proposes a structured action with resolved values.
• 4. The proposal is displayed for user approval.
• 5. Upon approval, the action is executed via the connector.
• 6. The proposal, approval, and result are logged in an audit trail.

Evidence

When you ask about something in your connected data, the reply comes back with the specific source it drew from attached to it.

Source

The model produces a proposal, a structured description of the exact action it wants to take, with the concrete arguments filled in.

Source