Build a broker copilot that actually helps close work

Last updated: March 2026

A good broker copilot does not just chat. It reads the right records, proposes the next action, and keeps the broker in control when external communication or data changes are involved.

Target Workflow

The minimal useful copilot usually answers four questions:
  1. What is the current state of the deal?
  2. Which lenders look like the best next fit?
  3. What has already been sent or discussed?
  4. What should the broker do next?

Core building blocks

Primary records
Deals, deal financials, deal properties, placements, term sheets, and account/team context.
AI-native workflow
Use MCP for interactive tooling or use a server-side orchestrator that can call the Lev API deterministically.
Decision layer
Let the model summarize trade-offs, but keep writes and outbound actions behind explicit approval.
Review layer
Show the retrieved records and the exact proposed action before the user confirms it.

Suggested Tool Sequence

  • Read the deal and its financial context.
  • Inspect recent placements and any available term sheets.
  • Trigger or read lender search results if the next best-fit lenders are not already known.
  • Summarize the opportunity and identify the next recommended broker action.
  • If a write is required, show the exact change and request confirmation.

Human Review Points

Keep the broker in charge
Use the model for synthesis and recommendation. Keep the broker responsible for final approval whenever a placement, status change, or outbound communication could affect a live relationship.
  • Review before creating or updating placements.
  • Review before summarizing or comparing term sheets for external sharing.
  • Review before any outbound lender communication is triggered by an agent workflow.

What's next

Use the underlying reference pages to implement the copilot's data access layer.

Continue to Deals