Frequently Asked Questions

General Questions

What is Thought Ledger?

Thought Ledger is a local-first AI decision memory system that helps you make better decisions by learning from your past choices. It runs entirely on your device, ensuring complete privacy and control over your data.

How is it different from other AI tools?

Unlike cloud-based AI services, Thought Ledger:

Never sends your data to the cloud
Works completely offline after initial setup
Learns from your personal decision patterns
Requires no API keys or subscriptions
Gives you full ownership of your AI memories

What does “decision memory” mean?

Decision memory is the accumulated knowledge of your decisions, their contexts, and outcomes. Thought Ledger helps you build a personal database of your choices that provides insights for future decisions.

Privacy & Security

Is my data really private?

Yes. All your data stays on your device:

No cloud uploads or synchronization
No third-party data sharing
Local processing only
You control all exports and backups

Can Thought Ledger work without internet?

Absolutely. Once installed, Thought Ledger works completely offline. The only time you need internet is for the initial download of AI models.

How is my data protected?

Local storage: All data stored in SQLite on your device
Encryption: Sensitive decisions can be encrypted
No telemetry: We don’t collect usage data or analytics
Open source: Code is transparent and auditable

Can someone access my decision data?

Only if they have physical access to your device and your files. We recommend:

Regular system security updates
Optional password protection for sensitive decisions
Encrypted backups if you store them externally

Technical Requirements

What are the minimum system requirements?

RAM: 8GB (16GB recommended)
Storage: 10GB free space
OS: macOS 11+, Ubuntu 20.04+, or equivalent
CPU: 4+ cores (8+ recommended)

Can it run on older hardware?

It depends on your RAM:

8GB RAM: Use smaller models (llama3.2:1b)
16GB RAM: Good performance with medium models
32GB+ RAM: Can run larger, more capable models

Does it work on Windows?

Yes, through WSL2 (Windows Subsystem for Linux). Performance is excellent on modern Windows hardware.

How much disk space do AI models need?

Model	Size	Use Case
llama3.2:1b	~1GB	Basic decisions, low RAM
llama3.2:3b	~2GB	Good balance of performance
llama3.2:8b	~5GB	Recommended for most users
llama3.1:70b	~40GB	Complex decisions, high RAM

Installation & Setup

How long does installation take?

Quick install: 5-10 minutes
Manual install: 10-15 minutes
AI model download: 2-10 minutes (depending on model size and internet speed)

Can I move Thought Ledger to another device?

Yes! Your decision database is portable:

Export your data from the old device
Install Thought Ledger on the new device
Import your data
Download your preferred AI models

Do I need to reinstall when updating?

No. Updates are handled through:

git pull origin main
npm install
npm run build

Your data and models remain intact.

Usage & Features

How accurate are the AI insights?

Insight quality improves with:

More decisions: Larger decision history
Better context: Detailed decision information
Regular use: Consistent pattern recognition
Model size: Larger models provide better analysis

Can Thought Ledger make decisions for me?

No. Thought Ledger is a decision assistant, not a decision maker. It provides:

Context from your past decisions
Pattern analysis and insights
Risk assessment based on your history
Blind spot identification

The final decision is always yours.

How many decisions can it store?

Technically unlimited, but practical limits depend on:

Storage space: ~1KB per decision with context
Performance: Best performance with <100K decisions
Search speed: Larger databases may be slower

Can I use it for team decisions?

While designed for personal use, you can:

Export/import decision databases
Manual sharing of specific decisions
Separate installations for team members

Future versions may include dedicated team features.

AI Models & Performance

Which AI model should I use?

llama3.2:1b: Best for older hardware, basic insights
llama3.2:3b: Good balance, recommended starting point
llama3.2:8b: Best insights, needs 16GB+ RAM
codegemma:7b: Specialized for technical decisions

Can I use different AI models?

Yes! You can:

Download multiple models: ollama pull <model-name>
Switch between them: Update your configuration
Use specialized models: For specific decision types

How do I improve AI performance?

Use larger models: If you have sufficient RAM
Provide better context: More detailed decision information
Regular usage: Builds better pattern recognition
Keep software updated: Latest optimizations and features

Data Management

How do I backup my decision data?

# Export your data
npm run export-data

# Manual backup
cp data/decisions.db backups/decisions-$(date +%Y%m%d).db

Can I export my decisions?

Yes, multiple formats available:

JSON: For data analysis and migration
CSV: For spreadsheet applications
Markdown: For documentation
PDF: For sharing and archiving

What happens if I lose my data?

If you don’t have backups:

Start fresh: New decision database
Rebuild gradually: New decisions will build new patterns
Learn from experience: Future decisions benefit from new data

Always maintain regular backups!

Troubleshooting

Thought Ledger feels slow, what can I do?

Check system resources: Ensure sufficient RAM and CPU
Use smaller model: If memory is limited
Clear cache: rm -rf .cache/ && npm start
Update software: Latest performance optimizations

AI responses seem generic, how to improve?

Add more context: Detailed decision information
Build decision history: More past decisions to learn from
Use larger model: If hardware permits
Provide feedback: Rate AI responses to improve quality

Ollama keeps failing, what to do?

Check system requirements: Ensure sufficient resources
Restart service: pkill -f ollama && ollama serve
Reinstall Ollama: If corruption is suspected
Try different model: Some models work better on certain hardware

Support & Community

How do I get help?

Documentation: Full docs
Troubleshooting: Common issues
GitHub Issues: Report bugs
Discussions: Ask questions

Can I contribute to Thought Ledger?

Yes! We welcome contributions:

Code contributions: Features and bug fixes
Documentation: Improve guides and docs
Testing: Report issues and suggest improvements
Community: Help others in discussions

Is Thought Ledger open source?

Yes. Thought Ledger is fully open source under the MIT license. You can:

View the source: Complete transparency
Modify for your needs: Custom installations
Contribute improvements: Help the project grow
Self-host: Complete control over your installation

Still have questions? Join our community or check the documentation for more information.