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

# Quickstart

In five steps, your first conversation becomes usable data.

Five steps: connect a call source, set up an agent, record, create an Insights report, then connect your tools. By the end, you have a conversation that is transcribed, analyzed and ready to query.

{% hint style="info" %}
Before you start: a Praiz account.
{% endhint %}

{% stepper %}
{% step %}

### Connect a conversation source

Praiz captures your calls on three channels. Connect the one you use.

{% tabs %}
{% tab title="Video meetings" %}
Connect your calendar. Praiz sends a bot into your upcoming Zoom, Google Meet and Microsoft Teams meetings to record them. See "Over video".
{% endtab %}

{% tab title="Phone" %}
Praiz pulls your calls from your phone system: Aircall, Ringover or Minari. Follow the integration steps for your tool. See "Over the phone".
{% endtab %}

{% tab title="In person" %}
For face-to-face meetings, download the Praiz mobile app (iOS and Android). See "In person".
{% endtab %}
{% endtabs %}

What to expect: at least one source connected, ready to record your next calls.
{% endstep %}

{% step %}

### Set up your first agent

An agent reads your conversations and extracts the information you choose: a summary, a follow-up email, objections, churn signals.

* Import a ready-made agent from the "Agent library", for example a summary or a call scoring agent. Then define which calls it applies to.
* Or build a custom agent that fits your context.

What to expect: an active agent that applies to your next conversations.
{% endstep %}

{% step %}

### Record your first call

Start a call with the source you just connected. In a video meeting, the bot asks to join. The host admits or refuses it. The host can also remove the bot at any point during the meeting.

A few seconds after the call ends, the transcript and your agents' output are available on the call page. Ask questions about that call with Ask Praiz.

What to expect: your first conversation is transcribed and analyzed. Use those first results to adjust your agents' instructions.
{% endstep %}

{% step %}

### Create an Insights report

Insights analyze one agent's data across all of your conversations. They surface trends: recurring objections, the most cited competitors, churn signals. The agent must have analyzed at least 3 calls.

Create a report from an agent, then schedule it (daily, weekly, monthly). See "Insights" and "Recurring Insights".

What to expect: a report sent at the frequency you chose.
{% endstep %}

{% step %}

### Connect your tools (optional)

* CRM: Praiz fills in your fields (HubSpot, Salesforce, Pipedrive and others). See "CRM sync".
* MCP server: query your whole conversational database in plain language from Claude, ChatGPT or Gemini. See "MCP server".
* Slack: receive your agents' results and your reports on your channels. See "Smart Actions".
  {% endstep %}
  {% endstepper %}

## Related pages

* For the vocabulary used here (agents, conversational database, Ask Praiz, Insights, MCP), see "Key concepts".