Skip to content

Start Here

If you are a first-time buyer or operator, this is the recommended first path.

Codeer helps an expert team bring judgment into customer conversations and workflow moments. Instead of building a scripted chatbot, you teach an Agent how your team clarifies requests, gives safe guidance, makes bounded decisions, knows when to hand off, and learns from real conversations.

In this getting started flow, you will build a fictional Consultation Desk Agent for an expert service business. We use intake as the hero example because it is easy to understand and test, but the same pattern also works for review, follow-up, and decision support later in the customer journey.

Success for this guide is simple: complete one full beginner loop from first draft to stakeholder feedback to a reusable test case.

Consultation Desk running in the current Agent Editor and Live Test layout

What You'll Build

By the end of this guide, you will have:

  • A first version of a Consultation Desk Agent created with Copilot
  • A published Web channel with a shareable slug-based URL
  • Stakeholder feedback collected from real conversations
  • One risky response turned into a reusable case in Test Suite

The Example Story

Your Consultation Desk Agent should:

  • Welcome people who are not sure which service or consultation they need
  • Ask 2 or 3 clarifying questions before deciding on a next step
  • Choose one of three outcomes: recommend the right consultation, suggest a safe next step, or hand off to a human
  • Avoid inventing services, prices, guarantees, or outcomes that your team has not approved

Explore other patterns after the hero example

After you finish the hero example, use Examples as the reference layer for other expert touchpoints:

Prerequisites

Before you begin, make sure you have:

  • A Codeer.ai account
  • Access to a workspace where you can edit Agents and publish channels
  • A modern browser such as Chrome, Firefox, or Edge
1Set Up Workspace
↓
2Build with Copilot
↓
3Test in Live Test
↓
4Publish to a Controlled Audience
↓
5Review Feedback in Histories
↓
6Add a Reusable Eval Case
↓
7Update the Agent and Run Test Suite

Step 1: Set Up a Workspace

Create a clean place for this pilot so your Agent, channel, and feedback stay together.

→ Set Up Your Workspace

Step 2: Build the First Version

Use Copilot to draft the first Consultation Desk Agent and define what it should and should not do as an expert touchpoint.

→ Build the First Version

Step 3: Test and Fix

Run one normal conversation and one risky conversation in Live Test, then use Copilot to tighten the behavior.

→ Test and Fix with Live Test

Step 4: Share for Feedback

Publish a Web channel, send it to a few stakeholders, and ask them to leave feedback on exact replies.

→ Share for Feedback

Step 5: Review Feedback, Save a Case, Fix the Agent, and Run the Test

Review what happened in Histories, ask Copilot what changed the behavior, use Add Case to save one important issue as an eval case, update the Agent in Editor, and run the saved case in Test Suite.

→ Feedback and Improve

Next Steps

After this first feedback loop, continue with: