> For the complete documentation index, see [llms.txt](https://polser.gitbook.io/polser-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://polser.gitbook.io/polser-docs/features/smart-responses.md).

# Smart Responses

{% @arcade/embed flowId="nVdU5N9dl4KyRB8F6TuN" url="<https://app.arcade.software/share/nVdU5N9dl4KyRB8F6TuN>" %}

## How it works

***

{% tabs %}
{% tab title="Topic Detection" %}
Polser identifies topics from incoming messages, triggering a Smart Response panel when confidence thresholds are met.

<mark style="background-color:yellow;">**Confidence thresholds (for advanced users ;-)**</mark>

* High: `intent.score > 0.8`
* Medium: `intent.score > 0.4`
* Low: `intent.score ≤ 0.4`
  {% endtab %}

{% tab title="Smart Response Panel" %}
Offers ready-made answers, verifications, warnings, and resources. Agents can tailor responses before sending.
{% endtab %}

{% tab title="Edit before sending" %}
Agents can edit specific answers or warning before sending them to their customers.
{% endtab %}

{% tab title="Send in another language" %}
In multilingual conversations, Smart Responses will be translated by default to the user’s language.
{% endtab %}

{% tab title="Customization" %}
Agents manage responses through the desktop admin panel. Customizable elements include name, confidence level, content, and integrations. Responses can also be temporarily disabled.
{% endtab %}

{% tab title="Find topics" %}
Agents can locate any Smart Response on-demand, regardless of the latest detected topic.
{% endtab %}
{% endtabs %}

***

{% hint style="success" %}
Smart Responses is available for both mobile and web applications within Polser, **but** **they can only be edited or created in the browser**.
{% endhint %}

***

<details>

<summary>Planned enhancements</summary>

Our future plans include:

* **More intelligence**: We currently detect the topic by analyzing the last message. We aim to maintain the context of previously sent messages for more accurate topic detection.
* **More integrations**: we plan to offer integrations with more third party tools such as Payments, Billing, CRM, Catalogs, etc
* **Intelligent Recommendations:** Polser will offer intelligent recommendations for edits to Smart Responses based on conversational analysis, usage frequency, and sentiment analysis of customer feedback.

</details>