> ## 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.

# Memory Categories

> How MemSync tags and labels memories for optimized retrieval

# Memory Categories

MemSync automatically tags memories with category labels, making it easier to find relevant context and build comprehensive user profiles. Understanding these categories helps you optimize memory retrieval and user personalization.

## Available Categories

MemSync uses 10 predefined categories that cover the most important aspects of user information:

<CardGroup cols={2}>
  <Card title="Identity" icon="id-card">
    Personal information including name, age, background, location, and origin
  </Card>

  <Card title="Career" icon="briefcase">
    Work history, profession, studies, and professional goals
  </Card>

  <Card title="Interests" icon="star">
    Hobbies, passions, interests, and recreational activities
  </Card>

  <Card title="Relationships" icon="users">
    Family, friends, social connections, and relationship patterns
  </Card>

  <Card title="Health" icon="heart-pulse">
    Medical history, fitness goals, wellness information, and health behaviors
  </Card>

  <Card title="Finance" icon="dollar-sign">
    Budget, investments, financial goals, income, and financial background
  </Card>

  <Card title="Learning" icon="graduation-cap">
    Educational background, skills, courses, and knowledge goals
  </Card>

  <Card title="Travel" icon="plane">
    Past trips, future travel plans, favorite destinations, and travel history
  </Card>

  <Card title="Productivity" icon="list-check">
    Tasks, projects, time management, and organizational systems
  </Card>

  <Card title="Private" icon="lock">
    Sensitive or confidential personal information
  </Card>
</CardGroup>

## Category Details

### Identity

Captures who the user is at their core - their fundamental personal information and background.

**Examples:**

* "Lives in New York City, originally from California"
* "29 years old, married with two children"
* "Identifies as a creative problem-solver and lifelong learner"

```json theme={null}
{
  "memory": "Lives in San Francisco, originally from Chicago",
  "categories": ["identity"],
  "type": "semantic"
}
```

### Career

Professional life, work experience, and career-related information.

**Examples:**

* "Works as a Senior Data Scientist at Microsoft"
* "Has 8 years of experience in machine learning"
* "Currently pursuing MBA to transition into product management"

```json theme={null}
{
  "memory": "Recently promoted to Team Lead at their tech startup",
  "categories": ["career"],
  "type": "episodic"
}
```

### Interests

Hobbies, passions, and activities the user enjoys in their personal time.

**Examples:**

* "Passionate about photography and hiking"
* "Enjoys cooking Italian cuisine on weekends"
* "Avid reader of science fiction novels"

```json theme={null}
{
  "memory": "Recently started learning guitar and practicing daily",
  "categories": ["interests", "learning"],
  "type": "episodic"
}
```

### Relationships

Social connections, family relationships, and interpersonal dynamics.

**Examples:**

* "Married to Sarah, together for 5 years"
* "Close relationship with parents, calls them weekly"
* "Mentors junior developers at work"

```json theme={null}
{
  "memory": "Has a close group of college friends who meet monthly for board games",
  "categories": ["relationships", "interests"],
  "type": "semantic"
}
```

### Health

Physical and mental health information, fitness goals, and wellness practices.

**Examples:**

* "Training for a marathon, runs 30 miles per week"
* "Follows a vegetarian diet for ethical reasons"
* "Practices meditation for stress management"

```json theme={null}
{
  "memory": "Recently started therapy to work on work-life balance",
  "categories": ["health", "productivity"],
  "type": "episodic"
}
```

### Finance

Financial situation, goals, and money-related decisions and preferences.

**Examples:**

* "Saving for a house down payment, budget of \$500K"
* "Invests in index funds and retirement accounts"
* "Budgets carefully and tracks all expenses"

```json theme={null}
{
  "memory": "Recently started investing in individual stocks after research",
  "categories": ["finance", "learning"],
  "type": "episodic"
}
```

### Learning

Educational pursuits, skill development, and knowledge acquisition.

**Examples:**

* "Currently learning Spanish through Duolingo"
* "Taking online courses in data visualization"
* "Reading books about behavioral psychology"

```json theme={null}
{
  "memory": "Completed AWS certification and planning to pursue advanced certification",
  "categories": ["learning", "career"],
  "type": "episodic"
}
```

### Travel

Travel experiences, plans, preferences, and location-related information.

**Examples:**

* "Loves exploring national parks, visited 15 so far"
* "Planning a trip to Japan for cherry blossom season"
* "Prefers sustainable travel and eco-friendly accommodations"

```json theme={null}
{
  "memory": "Recently returned from a solo backpacking trip through Southeast Asia",
  "categories": ["travel", "interests"],
  "type": "episodic"
}
```

### Productivity

Work habits, organizational systems, tools, and productivity preferences.

**Examples:**

* "Uses Notion for project management and note-taking"
* "Prefers working in focused 2-hour blocks"
* "Struggles with email management and time blocking"

```json theme={null}
{
  "memory": "Recently adopted Getting Things Done methodology for task management",
  "categories": ["productivity", "learning"],
  "type": "episodic"
}
```

### Private

Sensitive information that should be handled with special care.

**Examples:**

* Social Security numbers, passwords, private addresses
* Medical diagnoses or sensitive health information
* Financial account numbers or private financial details

```json theme={null}
{
  "memory": "Has been dealing with anxiety, takes medication as needed",
  "categories": ["private", "health"],
  "type": "semantic"
}
```

## Multiple Categories

Memories can belong to multiple categories when they span different aspects of a user's life:

### Common Combinations

<AccordionGroup>
  <Accordion title="Career + Learning">
    Professional development activities

    ```json theme={null}
    {
      "memory": "Taking a course in advanced React to improve frontend skills at work",
      "categories": ["career", "learning"],
      "type": "episodic"
    }
    ```
  </Accordion>

  <Accordion title="Interests + Health">
    Hobbies that also benefit physical or mental health

    ```json theme={null}
    {
      "memory": "Practices yoga 3 times a week for both fitness and stress relief",
      "categories": ["interests", "health"],
      "type": "semantic"
    }
    ```
  </Accordion>

  <Accordion title="Relationships + Travel">
    Social activities involving travel

    ```json theme={null}
    {
      "memory": "Planning annual ski trip with college friends to Colorado",
      "categories": ["relationships", "travel"],
      "type": "episodic"
    }
    ```
  </Accordion>

  <Accordion title="Productivity + Career">
    Work-related organizational systems

    ```json theme={null}
    {
      "memory": "Uses Slack and Asana to coordinate with remote team members",
      "categories": ["productivity", "career"],
      "type": "semantic"
    }
    ```
  </Accordion>
</AccordionGroup>

## Category-Based Search

Use categories to find specific types of information about users:

### Single Category Search

```python theme={null}
# Find career-related memories
career_memories = memsync.search_memories(
    query="What does the user do professionally?",
    categories=["career"],
    limit=10
)

# Find current interests
interest_memories = memsync.search_memories(
    query="What are the user's hobbies?",
    categories=["interests"],
    limit=8
)
```

### Multiple Category Search

```python theme={null}
# Find learning and career overlap
professional_development = memsync.search_memories(
    query="What skills is the user developing?",
    categories=["career", "learning"],
    limit=15
)

# Find health and productivity intersection
wellness_habits = memsync.search_memories(
    query="How does the user maintain work-life balance?",
    categories=["health", "productivity"],
    limit=10
)
```

### All Categories Search

```python theme={null}
# Comprehensive user overview
all_memories = memsync.search_memories(
    query="Tell me about this user",
    limit=20,
    rerank=True
)
```

## Category Distribution Insights

Understanding how memories are distributed across categories can provide insights:

### Typical Distribution

* **Career**: 25-30% of memories
* **Interests**: 20-25% of memories
* **Learning**: 15-20% of memories
* **Relationships**: 10-15% of memories
* **Health**: 8-12% of memories
* **Productivity**: 6-10% of memories
* **Identity**: 5-8% of memories
* **Travel**: 3-7% of memories
* **Finance**: 2-5% of memories
* **Private**: 1-3% of memories

<Note>
  Distribution varies significantly based on user conversation patterns, application type, and personal disclosure preferences.
</Note>

## Best Practices

### For Memory Extraction

<AccordionGroup>
  <Accordion title="Rich Context">
    Provide detailed conversation context to help MemSync accurately categorize memories.

    ```python theme={null}
    # Good: Provides context for categorization
    messages = [
        {"role": "user", "content": "I've been learning Python at work to automate our data pipeline tasks"},
        {"role": "assistant", "content": "That's great! Automation can save a lot of time in data workflows"}
    ]
    # Result: Likely categorized as ["career", "learning"]
    ```
  </Accordion>

  <Accordion title="Specific Information">
    Include specific details that help with accurate categorization.

    ```python theme={null}
    # Better than: "I like to exercise"
    # Use: "I go rock climbing twice a week at the local gym for fitness and stress relief"
    # Result: Categories ["interests", "health"]
    ```
  </Accordion>

  <Accordion title="Multiple Contexts">
    Don't avoid mentioning when activities span multiple life areas.

    ```python theme={null}
    # Good: Shows intersection of categories
    "I'm taking evening MBA classes while working full-time to prepare for a career transition"
    # Result: Categories ["learning", "career"]
    ```
  </Accordion>
</AccordionGroup>

### For Search Optimization

<AccordionGroup>
  <Accordion title="Category-Specific Queries">
    Tailor your search queries to specific categories for better results.

    ```python theme={null}
    # Instead of generic queries, be category-specific
    career_query = "What is the user's professional background and current role?"
    interest_query = "What activities does the user enjoy in their free time?"
    ```
  </Accordion>

  <Accordion title="Cross-Category Insights">
    Look for patterns across categories to understand user holistically.

    ```python theme={null}
    # Find connections between work and personal life
    work_life_balance = memsync.search_memories(
        query="How does the user balance professional and personal priorities?",
        categories=["career", "health", "relationships"],
        limit=15
    )
    ```
  </Accordion>

  <Accordion title="Progressive Refinement">
    Start broad, then narrow down based on initial results.

    ```python theme={null}
    # Step 1: Broad search
    general_memories = memsync.search_memories(query="user overview", limit=20)

    # Step 2: Focus on most relevant categories found
    specific_memories = memsync.search_memories(
        query="user career goals", 
        categories=["career", "learning"], 
        limit=10
    )
    ```
  </Accordion>
</AccordionGroup>

## Category Evolution

Categories can provide insights into how users change over time:

### Tracking Changes

* **New categories appearing**: User exploring new areas of life
* **Category frequency changes**: Shifting priorities or life phases
* **Cross-category connections**: Growing integration of different life aspects

### Example Evolution

```python theme={null}
# Early conversations: Heavy focus on career
early_categories = ["career", "learning", "productivity"]

# Later conversations: More balanced life interests
later_categories = ["career", "interests", "relationships", "health", "travel"]
```

## Next Steps

<CardGroup cols={2}>
  <Card title="Semantic Search" icon="magnifying-glass" href="/essentials/search">
    Learn how to search across categories effectively
  </Card>

  <Card title="User Profiles" icon="user" href="/essentials/profiles">
    Understand how categories combine to create user profiles
  </Card>

  <Card title="Memory Types" icon="brain" href="/essentials/memory-types">
    Learn about semantic vs episodic memory classification
  </Card>

  <Card title="API Reference" icon="code" href="/api-reference/introduction">
    Explore API endpoints for category-based operations
  </Card>
</CardGroup>