Skip to main content

​
What is User Memory?

User Memory is a secure, personal, and persistent storage system that maintains important information about your preferences, decisions, and key details across all interactions with your digital twin.

Secure & Private

All data is encrypted and private to your account only

Permanent Storage

Information persists across Digital Twins, sessions, weeks, and months

Selective Storage

Only important, relevant information is stored

AI-Powered

Your digital twin intelligently decides what to remember

You are in Control

You decide what is stored in your memory and manage your parameters

Works with Assistants

Use the set_in_memory and get_from_memory agents in your Assistant’s instructions to save/retrieve parameters in/from user memory.
Take complete control of your User Memory!You have full authority over every preference and parameter stored in your personal memory system, and the User Memory Optimizer assistant empowers you to efficiently manage, reorganize, and optimize this valuable data. Through intelligent namespace organization, standardized naming conventions, and storage optimization, you can transform scattered preferences into a streamlined, well-structured system that enhances AI personalization while reducing memory footprint—giving you faster responses, better assistance, and complete control over how your digital assistant understands and serves your unique needs. 🎯

​
How User Memory Works

1

Automatic Detection

Your digital twin identifies important information during conversations (preferences, goals, company details, etc.)
2

Manual Storage

Ask your digital twin to remember something about you or the conversation (ex: “Remember that I prefer decorator syntax for Javascript functions”)
3

Intelligent Storage

Key information is automatically stored in structured parameters with descriptions and organized by namespaces (personal, work, preferences, etc.)
4

Cross-Session Access

Stored information becomes available in all future conversations and Digital Twins, providing instant personalization
5

Continuous Learning

Your digital twin updates and refines stored information as it learns more about your needs and preferences

​
What Gets Stored in User Memory?

  • Personal Information
  • Project Context
  • Preferences & Style
  • Technical Preferences
  • Assistant Integration
  • Role & Position: Job title, responsibilities, expertise areas, seniority level
  • Company Details: Organization name, industry, team structure, department
  • Contact Information: Preferred communication methods, time zones, availability
  • Personal Preferences: Working style, communication preferences, personality traits
  • Background: Education, certifications, years of experience

​
Benefits of User Memory

Instant Context

No need to repeat basic information in new conversations—jump straight into productive work

Personalized Experience

Responses tailored to your specific needs, role, and communication style preferences

Faster Productivity

Jump straight into work without setup or explanation—your twin knows your context

Continuous Improvement

Experience gets better as your digital twin learns more about your patterns and needs

Consistent Quality

Maintains quality and style across all interactions, regardless of which assistant you use

Proactive Assistance

Anticipates needs based on stored preferences and suggests relevant solutions

Cross-Platform Continuity

Your preferences follow you across different Digital Twins and specialized assistants

Reduced Token Usage

Efficient storage reduces the need for lengthy context explanations in each conversation

Smart Recommendations

Provides suggestions aligned with your established preferences and past decisions

​
Managing Your User Memory

​
Viewing Stored Information

Your digital twin can display your current User Memory parameters in various formats: 
Ask: "What do you remember about me?" 
Ask: "Show my user memory in JSON format"
Ask: "Display my preferences as a table"
Ask: "What's in my work namespace?"
Example Output: User Memory as JSON

​
Adding Information Manually

You can explicitly ask your digital twin to remember specific information with context:
Examples:
  • “Remember that I prefer concise documentation with clear examples”
  • “Store that our project deadline is December 15th in the project_context namespace”
  • “Remember that I work in the healthcare industry and need HIPAA-compliant solutions”
  • “Save my preference for TypeScript over JavaScript for all new projects”

​
Updating Stored Information

Simply provide new information during conversations—your twin will automatically update relevant parameters: 
"Actually, my role changed to Senior Technical Writer"
"Update my company name to Acme Corporation"
"I now prefer Vue.js over React for frontend development"
"My communication style should be more formal for client interactions"
Add to USer Memory
Review the structure of the parameters saved into the user memory in the Agent Details

​
Organizing with Namespaces

User Memory uses namespaces to organize information logically:
  • personal_info: Basic details, contact preferences, background
  • work_context: Role, company, team, responsibilities
  • project_details: Current initiatives, deadlines, stakeholders
  • technical_preferences: Languages, frameworks, tools, methodologies
  • communication_style: Tone, format, detail level preferences
  • learning_preferences: How you like to receive information and feedback

​
Clearing User Memory

Use this carefully - clearing User Memory removes all personalization!
Ask: "Clear my user memory" (removes everything)
Ask: "Clear my work_context namespace" (removes only work-related info)
Ask: "Remove my old project preferences" (selective removal)

​
Optimization with User Memory Optimizer Assistant

Over time, your user memory accumulates scattered preferences from countless conversations, creating inefficiencies and forcing repetitive explanations across different assistants. The User Memory Optimizer consolidates these fragmented parameters into organized namespaces, creating a streamlined system that delivers consistent personalization, reduces token waste, and ensures your preferences follow you seamlessly across all Digital Twins—making every AI interaction faster, smarter, and truly tailored to you. User Memory Optimizer

​
What the User Memory Optimizer Can Do:

Memory Analysis

Analyze your current User Memory for gaps, redundancies, and optimization opportunities

Strategic Organization

Organize stored information into logical namespaces and eliminate duplicate parameters

Personalization Enhancement

Identify missing preferences that could significantly improve your AI experience

Memory Cleanup

Remove outdated, conflicting, or redundant information to keep memory efficient and accurate

Namespace Restructuring

Reorganize parameters into logical categories for better accessibility and management

Performance Optimization

Streamline memory structure to reduce token usage and improve response speed

​
When to Use the User Memory Optimizer:

  • Initial Setup: When starting with a new digital twin or after account creation
  • Periodic Review: Monthly or quarterly memory optimization sessions
  • Role Changes: When your job, company, or responsibilities change significantly
  • Performance Issues: If responses feel less personalized or inconsistent over time
  • Memory Bloat: When you notice scattered or conflicting preferences
  • Cross-Assistant Issues: When different assistants seem to have different understandings of your needs
User Memory Optimizer Immediate Results: User Memory Optimized

​
Advanced User Memory Features

​
Conditional Preferences

Store context-dependent preferences for different scenarios: 
"Remember: Use formal tone for client communications, casual for internal team discussions"
"Store: Prefer detailed explanations for new concepts, brief summaries for familiar topics"
"Save: Use Python for data analysis projects, JavaScript for web development"

​
Temporal Context

Include time-sensitive information with automatic updates: 
"Remember my current project runs until Q2 2025, then I'll focus on the mobile app initiative"
"Store that I'm learning React this quarter, so provide more detailed explanations for React concepts"

​
Hierarchical Preferences

Establish preference hierarchies for decision-making: 
"Priority order for documentation: Accuracy > Clarity > Brevity > Style"
"Technical stack preference: TypeScript > JavaScript > Python > Java"

​
Best Practices for User Memory

1

Be Explicit About Important Information

Clearly state preferences, goals, and key details you want remembered. Include context about why these preferences matter for better AI decision-making.
2

Provide Context for Decisions

Explain the reasoning behind important choices so your AI can make better recommendations and understand trade-offs.
3

Update Information Regularly

Keep your digital twin current about changes in roles, projects, or preferences. Set reminders to review and update quarterly.
4

Use Descriptive Namespaces

Organize information logically using clear namespace categories that make sense for your workflow and responsibilities.
5

Leverage the User Memory Optimizer

Run this specialized assistant monthly to consolidate scattered preferences into an organized, efficient knowledge system.
6

Test Cross-Assistant Consistency

Verify that your preferences work consistently across different Digital Twins and specialized assistants.
7

Your Privacy is Protected

Your memory parameters are completely private and secure—accessible only to you, never shared externally, and protected by enterprise-grade security. Store preferences, learning styles, and personal details with confidence knowing your data remains under your full control.

​
Getting Started with User Memory

Pro Tip: Start by sharing key information about yourself, your role, and your current projects. Your digital twin will automatically begin building your User Memory profile with intelligent namespace organization!

​
Quick Start Checklist:

  • Share your role, seniority level, and company information
  • Mention your current projects, goals, and key deadlines
  • State your communication and documentation preferences
  • Identify your technical stack and tool preferences
  • Set up important recurring tasks or standard processes
  • Specify your learning style and feedback preferences

​
Example Initial Setup:

"I'm a Senior Technical Writer at Acme Corp, leading API documentation for our fintech platform. 
I prefer concise, well-structured docs with practical examples and clear error handling. 
My current priority is documenting our payment API v2.0, due December 15th. 
I work closely with backend developers and need technical accuracy over marketing fluff.
I prefer TypeScript examples, REST over GraphQL, and always include authentication details.
For communication, keep responses focused and actionable—I value efficiency over lengthy explanations."
This single interaction creates organized memory parameters across multiple namespaces:
  • personal_info: Senior Technical Writer, efficiency-focused
  • work_context: Acme Corp, fintech platform, API documentation lead
  • project_details: Payment API v2.0, December 15th deadline
  • technical_preferences: TypeScript, REST APIs, authentication focus
  • communication_style: Concise, actionable, technical accuracy priority

​
Troubleshooting User Memory

Information Not Remembered

Symptoms: AI doesn’t recall previously stated preferences Solution: Explicitly ask your digital twin to remember specific details, or use the User Memory Optimizer to identify gaps in storage

Outdated Information

Symptoms: AI references old job titles, completed projects, or changed preferences Solution: Provide updates during conversations, or ask the User Memory Optimizer to help clean up obsolete data

Inconsistent Responses

Symptoms: Different assistants provide conflicting advice or ignore established preferences Solution: Review and optimize your User Memory with the User Memory Optimizer assistant for better cross-platform consistency

Conflicting Preferences

Symptoms: AI seems confused about your preferences or provides contradictory suggestions Solution: Use the User Memory Optimizer to identify and resolve conflicting parameters, establishing clear preference hierarchies

Memory Bloat

Symptoms: Slow responses or repetitive questions about basic preferences Solution: Run the User Memory Optimizer to consolidate redundant parameters and streamline your memory structure

Missing Context

Symptoms: AI asks for information you’ve provided before or doesn’t understand your work context Solution: Ensure important context is explicitly stored and consider using namespace organization for better accessibility

​
Maximize Your Digital Twin Experience

User Memory transforms your digital twin from a helpful tool into a personalized AI assistant that truly understands your needs, preferences, and goals. The more thoughtfully you build and maintain your User Memory, the more valuable and efficient your digital twin becomes over time.

​
Success Metrics:

  • Reduced Setup Time: Jump into productive work immediately without explaining context
  • Consistent Quality: Receive responses that match your style and preferences across all assistants
  • Proactive Assistance: Get relevant suggestions and recommendations based on your stored preferences
  • Cross-Platform Continuity: Experience seamless personalization whether using specialized assistants or general conversation
Remember: User Memory is an investment in your AI experience—the more thoughtfully you build it, the more valuable and efficient your digital twin becomes!
Next Steps:
  • Try the “Personalize My Experience” assistant to go through a series of 5 questions that will help you build a comprehensive user memory foundation!
  • Try the “User Memory Optimizer” assistant to analyze and enhance your current User Memory setup for maximum personalization and efficiency! 🚀
  • Explore specialized assistants knowing your preferences will follow you seamlessly across all interactions!