> ## Documentation Index
> Fetch the complete documentation index at: https://help.agentsy.pro/llms.txt
> Use this file to discover all available pages before exploring further.

# Create your first AI Experience

> Use the Agentsy AI Experience Builder to prepare your AI, then bring it to life in your branded Agentsy website.

You don't need to design an AI assistant from scratch. The [Agentsy AI Experience Builder](https://www.agentsy.pro/build-your-ai) helps you prepare the most important parts of your experience before you configure them in Agentsy.

This guide follows the simplest path: **build with AI → add the results to Agentsy → test → share**.

> [!NOTE]
>
> Already completed the Builder? Skip to [Step 3: Bring your experience into Agentsy](#3-bring-your-experience-into-agentsy). If you want an overview of the entire process, see [Quickstart](/getting-started/quickstart).

## 1. Let the AI Experience Builder design your assistant

Open [Build Your AI](https://www.agentsy.pro/build-your-ai) and follow **Step 1**.

1. Copy the **AI Experience Builder** prompt.
2. Paste it into ChatGPT, Claude, Grok, or your preferred AI assistant.
3. Describe your work and answer the questions it asks. If you've previously discussed your business with that AI, it may already have useful context.
4. Review the output and correct anything that doesn't reflect your expertise or offering.

The Builder helps you create two key resources:

| Builder output | What it's for |
| - | - |
| **System Instructions** | Defines your assistant's role, behavior, tone, boundaries, and approach to human handoff. |
| **Knowledge Sources Checklist** | Identifies the information and documents your assistant should use to answer questions reliably. |

You don't need to write your own system instructions or invent a knowledge structure from zero. The Builder does the initial planning with you.

> [!TIP]
>
> If an output feels too generic, ask the AI to refine it using your real methodology, typical client questions, and examples. Review the result before using it.

## 2. Let the Experience Page Builder prepare your website

On the same [Build Your AI](https://www.agentsy.pro/build-your-ai) page, continue to **Step 2 — Build Your Experience**.

Use the **Experience Page Builder** prompt with the System Instructions you prepared in Step 1. It helps draft:

* A headline and introduction for your branded AI website.
* Messaging about the audience you serve and the problems you help solve.
* Supporting website content that invites visitors to explore your expertise.
* Suggested brand colors and visual direction.

You can edit these recommendations. The goal is a website that sounds like you and gives visitors a clear reason to start a conversation.

> [!NOTE]
>
> The builders prepare content and configuration recommendations. They do **not** automatically create your Agentsy account, upload your documents, or publish your website.

## 3. Bring your experience into Agentsy

Once the builders have prepared your materials:

1. [Create your Agentsy account](/getting-started/create-your-account) or sign in.
2. Open your AI Experience's configuration in your workspace.
3. Add the **System Instructions** generated in Step 1.
4. Use the **Knowledge Sources Checklist** to choose and upload the actual documents your assistant needs.
5. Apply the **website copy and branding** prepared in Step 2.
6. Review the available visitor access options and, if appropriate, add a way to contact or book time with you.

That's the main setup. You don't need to redesign everything inside Agentsy—the builders have already helped you prepare the content.

> [!WARNING]
>
> A Knowledge Sources Checklist is a plan, not a knowledge base. Your assistant can only use source materials that are actually made available to it. Don't upload confidential or sensitive information without confirming that its use is appropriate.

For help with an individual setting, see [System Instructions](/create-and-configure/system-instructions), [Upload Documents](/knowledge-management/upload-documents), [Customize Your AI Website](/branding-and-website/customize-your-ai-website), and [Public and Private Access](/create-and-configure/public-and-private-access).

## 4. Test the result

Open the visitor-facing AI website and try a few realistic questions:

* **Introduction:** "What can you help me with?"
* **Expertise:** Ask about a topic covered by your methodology or documents.
* **Uncertainty:** Ask about something the assistant cannot verify.
* **Human connection:** "I'd like to speak with you. What's the next step?"

Check that the assistant represents your expertise accurately, stays within its boundaries, and offers a human conversation when appropriate—not at every opportunity.

If an answer needs improvement, update the relevant instructions or source document and test again. Also check the website on mobile and in a fresh browser session.

See [Testing Your Assistant](/create-and-configure/testing-your-assistant).

## 5. Share your AI Experience

Once the first version works well, verify its access settings and share the branded website with your audience.

Start with people who would genuinely benefit from exploring your expertise. Explain **what they can discover through the experience**, rather than only announcing that you have an AI assistant.

See [Publish Your AI Experience](/share-and-grow/publish-your-ai-experience) and [Share Your AI Website](/share-and-grow/share-your-ai-website).

## First-experience checklist

* [ ] Completed the AI Experience Builder and reviewed its System Instructions.
* [ ] Reviewed the Knowledge Sources Checklist.
* [ ] Completed the Experience Page Builder and reviewed its website recommendations.
* [ ] Added the instructions, actual knowledge documents, and website content in Agentsy.
* [ ] Checked visitor access and any human contact option.
* [ ] Tested common questions, unknown questions, and the mobile experience.
* [ ] Confirmed the visitor-facing link is ready to share.

**Ready to start?** [Open the AI Experience Builder →](https://www.agentsy.pro/build-your-ai)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.