Skip to main content
ChatSyncs’ AI system is built around named agents — an Order Status Agent, a Sales Advisor Agent, a WhatsApp Marketing Agent, and so on — each with its own instructions, knowledge, and restrictions. Agent Routing then decides which agent answers a given customer message.

1. How to Create Your First AI Agent

Turn on AI Agents, build a named agent (name, description, system prompt, knowledge), and switch it on for your bot.

2. How to Train Your AI Chatbot

The full reference for every Knowledge Source type — Content, URL, File, Media, Google Sheet, and API.

3. Agent Routing

Decide which agent replies to which kind of customer message.

4. AI Agent Actions

The ##action## syntax that lets an agent do things, not just talk — assign labels, look up orders, hand off to a human.

5. How Do I Restrict AI on Specific Keywords?

Stop an agent from answering topics it shouldn’t touch.

6. How Do I Auto-Assign Conversations to a Department?

Hand a conversation to a human team instead of leaving it with the AI.

7. How Do I Build a Multi-Agent AI Workforce?

A worked example: three specialized agents working together on one bot.
Want the AI silent during business hours instead of always-on? See Disable AI Agent under How Do I Set Up Business Hours for My Bot?.

Frequently asked

How to Create Your First AI Agent — it walks through turning the system on, building your first named agent, and switching it on.
No — a single agent with a broad System Prompt and one Knowledge Set works fine for most bots. Multiple agents are worth it once different types of questions need very different instructions or actions — see How Do I Build a Multi-Agent AI Workforce?.
Yes — that’s a separate setting from anything here. See Disable AI Agent under How Do I Set Up Business Hours for My Bot?.

Learn From a Use Case

The linked use cases below were written against ChatSyncs’ earlier single-agent Intent Detection system. The screens they show (AI Agents and Intent Detection, Create Intent Detection) have since been replaced by the multi-agent system in this section — the underlying ideas (routing a message to the right handler) still apply, but the exact steps and fields will look different in the current UI. Use Agent Routing and AI Agent Actions for the current equivalent.