> ## Documentation Index
> Fetch the complete documentation index at: https://docs.memsync.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat Integration

> Import chat histories and conversations to extract memories

Import chat histories from various platforms to automatically extract and store conversational memories.

<RequestExample>
  ```bash cURL theme={null}
  curl -X POST 'https://api.memchat.io/integrations/chat' \
    -H 'X-API-Key: YOUR_API_KEY' \
    -H 'Content-Type: application/json' \
    -d '{
      "platform": "chatgpt",
      "chat_data": [
        {
          "timestamp": "2024-01-15T14:30:00Z",
          "role": "user",
          "content": "Tell me about machine learning applications in healthcare"
        },
        {
          "timestamp": "2024-01-15T14:30:15Z", 
          "role": "assistant",
          "content": "Machine learning has numerous applications in healthcare..."
        }
      ],
      "extract_mode": "conversations"
    }'
  ```

  ```python Python theme={null}
  import requests

  url = "https://api.memchat.io/integrations/chat"
  headers = {
      "X-API-Key": "YOUR_API_KEY",
      "Content-Type": "application/json"
  }

  data = {
      "platform": "chatgpt",
      "chat_data": [
          {
              "timestamp": "2024-01-15T14:30:00Z",
              "role": "user", 
              "content": "Tell me about machine learning applications in healthcare"
          },
          {
              "timestamp": "2024-01-15T14:30:15Z",
              "role": "assistant",
              "content": "Machine learning has numerous applications in healthcare..."
          }
      ],
      "extract_mode": "conversations"
  }

  response = requests.post(url, headers=headers, json=data)
  result = response.json()
  ```

  ```javascript JavaScript theme={null}
  const response = await fetch('https://api.memchat.io/integrations/chat', {
    method: 'POST',
    headers: {
      'X-API-Key': 'YOUR_API_KEY',
      'Content-Type': 'application/json',
    },
    body: JSON.stringify({
      platform: 'chatgpt',
      chat_data: [
        {
          timestamp: '2024-01-15T14:30:00Z',
          role: 'user',
          content: 'Tell me about machine learning applications in healthcare'
        },
        {
          timestamp: '2024-01-15T14:30:15Z',
          role: 'assistant', 
          content: 'Machine learning has numerous applications in healthcare...'
        }
      ],
      extract_mode: 'conversations'
    }),
  });

  const result = await response.json();
  ```
</RequestExample>

## Request Body

<ParamField body="platform" type="string" required>
  Chat platform source. Supported: `chatgpt`, `claude`, `telegram`, `whatsapp`, `slack`, `discord`
</ParamField>

<ParamField body="chat_data" type="array" required>
  Array of chat messages

  <Expandable title="Message Properties">
    <ParamField body="timestamp" type="string" required>
      ISO 8601 timestamp of the message
    </ParamField>

    <ParamField body="role" type="string" required>
      Message role: `user`, `assistant`, `system`, or participant name
    </ParamField>

    <ParamField body="content" type="string" required>
      Message content/text
    </ParamField>

    <ParamField body="message_id" type="string">
      Unique identifier for the message
    </ParamField>

    <ParamField body="metadata" type="object">
      Additional message metadata
    </ParamField>
  </Expandable>
</ParamField>

<ParamField body="extract_mode" type="string" default="conversations">
  Extraction strategy. Options: `conversations`, `insights`, `topics`, `all`
</ParamField>

<ParamField body="conversation_settings" type="object">
  Settings for conversation processing

  <Expandable title="Conversation Settings">
    <ParamField body="min_length" type="number" default="50">
      Minimum message length to consider
    </ParamField>

    <ParamField body="group_by_topic" type="boolean" default="true">
      Whether to group related messages by topic
    </ParamField>

    <ParamField body="include_system_messages" type="boolean" default="false">
      Whether to include system messages
    </ParamField>

    <ParamField body="time_gap_threshold" type="number" default="3600">
      Seconds gap to separate conversations
    </ParamField>
  </Expandable>
</ParamField>

<ParamField body="auto_categorize" type="boolean" default="true">
  Whether to automatically categorize extracted memories
</ParamField>

## Response

<ResponseField name="integration_id" type="string">
  Unique identifier for the chat integration
</ResponseField>

<ResponseField name="platform" type="string">
  Chat platform that was processed
</ResponseField>

<ResponseField name="status" type="string">
  Processing status: `processing`, `completed`, `error`
</ResponseField>

<ResponseField name="processing_stats" type="object">
  Statistics about the processing

  <Expandable title="Processing Statistics">
    <ResponseField name="total_messages" type="number">
      Total number of messages processed
    </ResponseField>

    <ResponseField name="conversations_identified" type="number">
      Number of distinct conversations found
    </ResponseField>

    <ResponseField name="memories_extracted" type="number">
      Number of memories created
    </ResponseField>

    <ResponseField name="topics_detected" type="array">
      Main topics identified in conversations
    </ResponseField>

    <ResponseField name="date_range" type="object">
      Date range of the chat history
    </ResponseField>
  </Expandable>
</ResponseField>

<ResponseField name="memories" type="array">
  List of extracted memories

  <Expandable title="Memory Properties">
    <ResponseField name="memory_id" type="string">
      Unique identifier for the memory
    </ResponseField>

    <ResponseField name="content" type="string">
      Extracted conversation content
    </ResponseField>

    <ResponseField name="category" type="string">
      Assigned category
    </ResponseField>

    <ResponseField name="conversation_topic" type="string">
      Main topic of the conversation
    </ResponseField>

    <ResponseField name="participants" type="array">
      Participants in the conversation
    </ResponseField>

    <ResponseField name="start_time" type="string">
      When the conversation started
    </ResponseField>

    <ResponseField name="end_time" type="string">
      When the conversation ended
    </ResponseField>
  </Expandable>
</ResponseField>

<ResponseExample>
  ```json Response theme={null}
  {
    "integration_id": "chat_int_789abc",
    "platform": "chatgpt",
    "status": "completed",
    "processing_stats": {
      "total_messages": 24,
      "conversations_identified": 3,
      "memories_extracted": 5,
      "topics_detected": [
        "Machine Learning in Healthcare",
        "Python Programming",
        "Career Development"
      ],
      "date_range": {
        "start": "2024-01-10T09:00:00Z",
        "end": "2024-01-15T17:30:00Z"
      }
    },
    "memories": [
      {
        "memory_id": "mem_chat_001",
        "content": "Discussed machine learning applications in medical diagnosis, specifically how neural networks can improve accuracy in identifying diseases from medical imaging. Learned about specific techniques and real-world implementations.",
        "category": "learning",
        "conversation_topic": "Machine Learning in Healthcare",
        "participants": ["user", "assistant"],
        "start_time": "2024-01-15T14:30:00Z",
        "end_time": "2024-01-15T14:45:00Z"
      }
    ]
  }
  ```
</ResponseExample>

## Extraction Modes

<Accordion title="Conversations">
  Groups related messages into conversation memories, preserving context and flow.
</Accordion>

<Accordion title="Insights">
  Extracts key insights, learnings, and important information from conversations.
</Accordion>

<Accordion title="Topics">
  Creates memories organized by topics discussed across conversations.
</Accordion>

<Accordion title="All">
  Combines all extraction modes for comprehensive memory creation.
</Accordion>

## Supported Platforms

<Accordion title="ChatGPT">
  * Export conversations from ChatGPT
  * Identifies learning topics and insights
  * Categories: primarily `learning`, `productivity`
</Accordion>

<Accordion title="Claude">
  * Claude conversation exports
  * Focus on analytical discussions
  * Categories: `learning`, `productivity`, `interests`
</Accordion>

<Accordion title="Telegram/WhatsApp">
  * Personal and group conversations
  * Relationship and social memories
  * Categories: `relationships`, `interests`, `travel`
</Accordion>

<Accordion title="Slack/Discord">
  * Professional and community discussions
  * Work-related and interest-based memories
  * Categories: `career`, `relationships`, `interests`
</Accordion>

## Privacy and Security

* **Data Processing**: Chat data is processed securely and not stored permanently
* **Participant Privacy**: Names and identifiers can be anonymized
* **Content Filtering**: Sensitive content is automatically filtered out
* **Retention**: Original chat data is deleted after processing