After this lesson, you will be able to: Design a conversational AI agent in Voiceflow with intents, knowledge bases, and a deployable chat or voice interface.
Voiceflow is purpose-built for conversation design, chatbots, voice agents, IVR. Where n8n/Zapier/Make think in workflows, Voiceflow thinks in conversations: blocks, branches, intents, and slots. This lesson builds a deployable customer support bot.
In Voiceflow, each block is a step in a dialog: a message, a question, an AI response, an API call, a branch. You connect blocks like a flowchart of what the bot says next. Intents (user goals) and entities (data extracted from input) drive the branching.
Diagram coming soon!
Voiceflow canvas: Start → Welcome message → Capture intent → branches into FAQ block, Pricing block, Human handoff block
1. Sign up at voiceflow.com (free tier).
2. Create a new chat assistant (not voice, same builder, different deployment).
3. Drag a 'Message' block: 'Hi! I'm a support bot. What can I help with?'
4. Drag a 'Choice' block. Add intents: 'Pricing', 'Bug Report', 'Other'.
5. From 'Pricing' branch: Message block with your pricing. From 'Bug Report': Capture block (asks for description), then API block to your ticket system, then Message confirming.
6. From 'Other': drag an 'AI' block. System prompt: 'You're a friendly support bot. Answer briefly.'
7. Click 'Run' (top right). Test in the inline chat.
1. Click 'Publish' (top right).
2. Choose 'Web Chat' integration. Voiceflow gives you an embed snippet.
3. Paste the snippet into any HTML page (an in-browser editor page is great for testing).
4. The bot is now live on a real webpage.
Same canvas, deploy to Alexa, Google Assistant, or a phone number via Twilio Voice. Your conversation logic doesn't change. Voiceflow handles the speech-to-text and text-to-speech layer.
Sign in and purchase access to unlock this lesson.