Create your SynTwin

This tutorial creates a SynTwin you can use to scale your daily communication.

You will create the SynTwin profile, add knowledge, create a chat deployment, then optionally add voice or video.

If you want the fastest sales path, use First sales loop instead.

1. Fill the SynTwin profile

Open the profile screen under Setup.

Add:

• Name — your name as visitors should see it.
• Role — your current role or sales function.
• Company — the company your SynTwin represents.
• Bio — what you do, who you help, and what expertise the SynTwin should reflect.
• Languages — languages your SynTwin can use.

Save the profile before continuing.

2. Add global sources

Open Train > Sources and upload owner-level material your SynTwin can use across deployments.

Good first sources:

• Product one-pagers
• Sales decks
• Pricing or packaging notes
• FAQs
• Discovery call transcripts
• Email examples
• Meeting notes

Supported upload types are .pdf, .docx, .pptx, .md, .txt, and .csv.

3. Add conversation style

Upload real communication examples: call transcripts, emails, meeting notes, or raw notes.

SynTwin uses them for two things:

• Knowledge: facts, product details, objections, customer language.
• Style: how you explain, ask, qualify, and follow up.

4. Create a chat deployment

In the Deploy sidebar group, create a deployment.

Use a goal like:

Answer visitor questions about Product XYZ and get the user to submit their email for follow-up conversations.

Then open the deployment settings and add instructions:

Make a very short introduction so the visitor can start speaking immediately.
Ask for their name, industry, and company so you can give a domain-specific introduction.

Choose channels:

• Chat works without voice or video setup.

In Sources, mark the global sources this deployment may use as Mentioned or Internal. Unselected sources are unavailable to the deployment.

5. Share and test

Open the deployment dashboard.

Use Try to test chat.

Use Share to copy a direct link, create a personalized link, or copy the embed snippet.

Before sharing widely, run a short readiness check:

• The intro is short.
• The SynTwin can answer the top three expected questions.
• The SynTwin asks for the visitor details you need.
• The deployment uses only relevant sources.
• Internal material is not marked Mentioned.
• The goal result feels accurate on a test session.

Your SynTwin is now ready for first real conversations.

6. Add voice or video later

Open Setup > Appearance when chat is working.

Add at least one avatar image.

For audio deployments, create a voice sample:

1. Click Clone voice.
2. Record a sample for a profile language.
3. Wait until it is ready.
4. Select the ready voice for that language.

For video deployments, add a video layer:

• Use a photo for an instant video layer.
• Record a video for a higher-quality clone.

The guided recording includes consent, speaking, and silent listening.

Then enable Audio or Video in the deployment.

7. Run your first human + SynTwin loop

Use the deployment in a real sales flow:

1. Send a personalized link to one visitor.
2. Let the SynTwin answer first questions and capture email.
3. Review the session summary before your human meeting.
4. Add meeting notes to the person afterward.
5. Use a follow-up link for later questions.
6. Answer reflection questions when the SynTwin gets stuck.

See Human + SynTwin workflow for the full loop.