Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.poly.ai/llms.txt

Use this file to discover all available pages before exploring further.

Level 1: Build your first agent. By the end of these 6 short lessons, you’ll have a real agent that answers questions, speaks with a voice you’ve chosen, and is safely versioned. No code required – estimated time: 30 minutes.

​
What you’ll build

You’ll go from an empty project to a working agent that:
  • Answers FAQ-style questions in both Chat and Call
  • Speaks with a voice and personality you’ve configured
  • Lives safely in versioned environments (Sandbox → Pre-release → Live)

​
Your 6-lesson journey

1

Create your project

Pick a name, set a language, and get your workspace ready
2

Give your agent a personality

Define who the agent is, how it speaks, and what it should never do
3

Teach it something

Add a simple FAQ topic so your agent can answer a real question
4

Choose a voice

Pick how your agent sounds and tune it for clarity
5

Learn environments

Understand how to test safely before anything goes live
6

Review what happened

Look at real conversations and confirm your agent did the right thing

​
Lessons

1. Create a project

Set up your first agent with a clear name and configuration

2. Edit agent behavior

Define personality, rules, and what the agent can and can’t do

3. Add a simple topic

Create a one-turn FAQ topic that works in Chat and Call

4. Edit voice settings

Choose how your agent sounds and adjust clarity

5. Environments

Test and promote changes safely: Sandbox → Pre-release → Live

6. Conversation reviews

Review real transcripts and check what your agent actually did
You’ve completed Level 1 when:
  • Your project is created and configured
  • Your agent has a personality and rules
  • A simple FAQ topic works in both Chat and Call
  • You’ve published and promoted through environments
  • You can find and review a conversation
Ready to level up? In Level 2, you’ll make your agent smarter – multi-turn topics, functions, response controls, audio tuning, and advanced diagnostics.
Last modified on April 30, 2026