> For the complete documentation index, see [llms.txt](https://help.botpenguin.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.botpenguin.com/ai-agents/creating-ai-agents-in-botpenguin/setting-up-your-ai-agent/meta-data.md). # Meta Data ### What is Meta Data? The Meta Data section lets you define which attributes your AI Agent should extract and track during a conversation. These attributes help the agent filter, analyze, and respond based on user-specific inputs like dates, locations, or preferences. *** ### When should you use this? Use this feature when: * You want your AI Agent to automatically capture specific details a user shares mid-conversation * You need conversation data structured for filtering, search, or analytics later * You're feeding captured data into a downstream workflow, like lead qualification or appointment scheduling *** ### Benefits of Meta Data * 🚀 Enables your AI Agent to capture and store critical user details automatically * 🎯 Powers contextual personalization in real-time conversations * 🔄 Enhances backend data filtering and segmentation for analytics or automation * 📈 Streamlines downstream workflows like appointment scheduling or lead qualification *** ### Expected Outcome After setting this up: * Your AI Agent will extract the attributes you've enabled, directly from the conversation * You will be able to search, filter, and analyze conversations using this captured data * Downstream workflows relying on this data will receive it automatically, without manual entry {% hint style="info" icon="lightbulb" %} Tip: This is how your setup should function once completed — every attribute marked for extraction gets pulled from the conversation automatically, with no manual tagging needed. {% endhint %} *** ### How to Use Meta Data 1. Go to BotPenguin and log in to your account. Go to "AI Agents" from the side panel and select your AI Agent.
2. Open Agent Personality & Instructions and go to the "Meta Data" tab. 3. You'll see a list of attributes the agent can extract. Each row shows: 1. Attribute Name — the label used to identify the data point 2. Category — Custom or System 3. Extract Meta Data — a toggle to turn automatic extraction on or off 4. Actions — a delete button to remove the attribute from the agent's scope {% hint style="info" %} If no metadata attributes have been added yet, you'll see a **"No Metadata Found"** empty state with a warning icon instead of the table. It tells you: *"You can create metadata attributes to extract from your data."* Click **"Manage Attribute"** to add your first one. {% endhint %}
4. Use the search bar to find a specific attribute by name, or filter by status: All, Active, or Non-active. 5. To create or modify attributes, click "Manage Attribute." This redirects you to the Custom Attributes section under BotPenguin's global Settings.
6. From there, define new attributes, assign types (e.g., Text, Date, Boolean), and configure options like mandatory fields or value validation. Refer to the full Managing Custom Attribute guide for detailed steps.
*** ### Best Practices > 💡 Only mark attributes for extraction if you'll actually use the captured data downstream — unused extraction adds noise, not value. > ⚠️ Attribute types (Text, Date, Boolean) should match the kind of answer you expect, or extraction accuracy drops. > 🚀 Use the Active/Non-active filter to periodically clean up attributes you're no longer using. > 🎯 Align the attributes you extract with what your downstream workflow actually needs — lead qualification and appointment scheduling often need different fields. *** ### Troubleshooting
Issue: The Meta Data tab shows "No Metadata Found" even though I expect attributes to exist. **Solution:** This means no metadata attributes have been created for this agent yet. Click "Manage Attribute" to create your first one under Custom Attributes.
Issue: An attribute isn't being extracted during conversations. **Solution:** Confirm the "Extract Meta Data" toggle is turned on for that attribute, and that its type (Text, Date, Boolean) matches the kind of input the user is expected to provide.
Issue: I can't find an attribute I created earlier. **Solution:** Use the search bar to search by name, and check the status filter — it may be set to "Active" or "Non-active" only, hiding the attribute you're looking for.
*** ### FAQs
What's the difference between Custom and System attributes? Custom attributes are ones you've created yourself under Custom Attributes. System attributes are built into BotPenguin by default.
Where do I actually create a new attribute? Click "Manage Attribute" from the Meta Data tab — this redirects you to the Custom Attributes section under BotPenguin's global Settings, where attributes are defined and typed.
What happens if I delete an attribute from the agent's scope? The agent stops extracting that attribute going forward. Previously captured data isn't affected.
Can I extract more than one attribute at a time? Yes, you can enable the "Extract Meta Data" toggle for as many attributes as you need.
What do I see if no metadata attributes exist yet? You'll see a "No Metadata Found" state with a warning icon and the message "You can create metadata attributes to extract from your data," instead of the attributes table.
*** ### Related Articles {% content-ref url="/pages/6ttpiKsmzt3DBoquI8zO" %} [Managing Custom Attribute](/settings/managing-custom-attribute.md) {% endcontent-ref %} {% content-ref url="/pages/cyflLQ2K7fe2GtP7lst9" %} [Agent Profile Setup: Core Instructions & Best Practices](/ai-agents/creating-ai-agents-in-botpenguin/setting-up-your-ai-agent/agent-profile-setup-core-instructions-and-best-practices.md) {% endcontent-ref %} {% content-ref url="/pages/cBMCtFeeMa2cfmq8H2Xx" %} [Training your AI Agent](/ai-agents/creating-ai-agents-in-botpenguin/training-your-ai-agent.md) {% endcontent-ref %} *** ### Support If your metadata isn't being extracted as expected, or attributes aren't showing correctly, reach out to us at . Our team will assist you within 48 business hours. *** ### 🎉 Completion 🎉 Congratulations, you have successfully completed this section! You're one step closer to mastering BotPenguin.