How Will the AI Use My Data?
Smart synthesis with complete privacy — how Mind Dump's AI stack works, and why your thoughts stay yours.
The AI Integration: Smart Synthesis, Complete Privacy
Mind Dump has built-in artificial intelligence to help you manage and navigate your Zettelkasten. Our AI tools are not meant to replace your thoughts, but rather to surface connections, generate embeddings, and help you synthesise your knowledge more efficiently.
What Mind Dump Uses: Our AI Stack
Two models through the Google Gemini ecosystem.
Mind Dump utilises 2 different models, through the Google Gemini ecosystem, to provide a fast responsive experience:
Gemini-flash-2.5-lite
This is our primary large language model. It basically acts as a quick-thinking friend, guiding you through analysing your notes, suggesting cross-references, and helping to categorise complex ideas in your Zettelkasten system.
Gemini embed
This model is vector embedding specific. It converts your notes' text into mathematical vectors so Mind Dump can understand the semantic meaning and context of what you write. This powers our advanced search and automatically surfaces related "slip-box" notes even if they don't share identical keywords.
The Industry Problem: Training & Human Review
Why most consumer AI tools are a poor fit for personal knowledge.
Many people are understandably hesitant to input personal thoughts, journals or proprietary knowledge bases into AI applications. There are two common ways that data privacy is compromised with consumer-facing AI tools (and the free-tier developer platforms):
1. Model training
Your prompts and notes go as feedback to the AI system. The provider then uses your data to train, enhance and improve future versions of their machine learning models. This puts your private information at risk of being memorised and possibly surfaced to other users elsewhere.
2. Human review
Random samples of user prompts and generated outputs are given to human reviewers who manually annotate and analyse the quality of the AI's responses.
This standard approach is totally unacceptable for a personal knowledge management tool like Mind Dump.
No Data Leakage — The Mind Dump Standard
Enterprise-grade API architecture with strict data covenants.
We believe your Zettelkasten should be a complete private extension of your mind. That's why Mind Dump is built on an independent Enterprise-Grade API Architecture.
Mind Dump exists in a dedicated, commercially secured developer tier, so you are protected by strict corporate data covenants in your interactions with the AI.
No training of the model
Google is prohibited by contract from using any text, prompts, system instructions, or notes processed through Mind Dump to train or improve its AI models. Your data is entirely your own.
No human supervision
Your inputs and outputs do not go through the usual public review queues. Any data you process using our AI features will never be seen by human eyeballs.
Strong boundary isolation
All API requests are processed within a secure cloud perimeter. When the model returns your answer, the server forgets the interaction, ensuring that your digital "Second Brain" is completely private.
Mind Dump lets you use the AI features knowing that your thoughts, theories and notes are totally private, secure and under your complete control.
Explore More
Learn how Mind Dump protects your data at rest, in transit, and in your most sensitive secret fields.