> For the complete documentation index, see [llms.txt](https://goldenrules.parloa.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://goldenrules.parloa.com/the-rules/speech-assets.md).

# Speech Assets

Best practice tips for optimizing your speech model

## <mark style="color:blue;">#1</mark> Use tellSlot instead of tellA and tellB

To make your speech model as stable as possible to accommodate variable content, it is recommended to use a single intent for standardized content from slots.

The background to this is also that utterances make little to no difference as to whether users provide, for example, a date of birth, a claim date, or a customer number.

This means:

{% hint style="danger" %}
**Don’t** create several intents for slot informations

tellBirthdate

tellInsuranceType

tellInsuranceNo
{% endhint %}

{% hint style="success" %}
**Do** create a single intent and use it everywhere you want the user to tell variable (slot) informations:
{% endhint %}

![](https://1775763192-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FTb7k9AiD3kgCg2gqzq7R%2Fuploads%2F5LLiLqz3vhO1SRcQnd5o%2FtellSlot-intent.png?alt=media\&token=6d467542-a9e3-4a45-9264-b035d7c55c66)