AI Nodes

Input Data ──▶ Prompt Template ──▶ AI Model ──▶ Parsed Output
                    │                  │
                    │                  ▼
              Liquid templates    Low/Medium/High
              + context data      model tier

messages: [
  { role: "system", content: "You are a helpful assistant..." },
  { role: "user", content: "Please analyze: {{ json.transcript }}" }
]

{{ json.meeting.title }}              // Access upstream data
{{ trigger.meetingPlanId }}           // Access trigger data
{{ json.attendees | size }}           // Use filters
{% for item in json.items %}          // Loops
  - {{ item.name }}
{% endfor %}

You are an expert meeting analyst. Your task is to:
1. Extract key decisions made
2. Identify action items with owners
3. Summarize the overall sentiment

Be concise and factual. Do not include information not present in the transcript.

Meeting: {{ json.meeting.title }}
Date: {{ json.meeting.startTime | date: "%B %d, %Y" }}
Duration: {{ json.callRecording.duration | divided_by: 60 }} minutes

Attendees:
{% for a in json.meeting.attendees %}
- {{ a.name }} ({{ a.email }})
{% endfor %}

Transcript:
{{ json.callRecording.transcript }}

Respond with a JSON object containing:
- "summary": string (2-3 sentences)
- "decisions": array of strings
- "action_items": array of { "task": string, "owner": string }
- "sentiment": "positive" | "neutral" | "negative"

{% if json.callRecording and json.callRecording.transcript != "" %}
  {{ json.callRecording.transcript }}
{% else %}
  No transcript available. Meeting title: {{ json.meeting.title }}
{% endif %}

[Load Meeting] ──▶ [AI Prompt: Summarize] ──▶ [Slack Post]

                              ┌──▶ [High Priority Handler]
[AI Prompt: Classify] ──▶ [If] ├──▶ [Medium Priority Handler]
                              └──▶ [Low Priority Handler]

[AI Agent: Research competitors] ──▶ [AI Prompt: Format report] ──▶ [Email]

[AI Prompt]
    ├── Success ──▶ [Continue]
    └── Error ────▶ [Fallback] ──▶ [Sink]