Privacy-first local AI memory system for storing and retrieving personal knowledge and documents
Project description
🧠 LocalMind
LocalMind is a privacy-first local AI memory system that lets you store, search, and retrieve personal knowledge and documents using AI.
Unlike typical document chat tools, LocalMind allows you to store personal information, notes, and files and retrieve them intelligently using hybrid search and local language models.
All data stays fully on your machine.
✨ Features
• Store personal knowledge locally
• Ask questions over your stored information
• Upload and search documents
• Hybrid retrieval (FAISS + BM25)
• Cross-encoder reranking for better answers
• CLI interface for quick access
• Privacy-first design (no cloud required)
📦 Installation
Install LocalMind from PyPI:
pip install localmind
⚙️ Windows Setup (If localmind command not found)
After installing, if you see:
localmind: The term 'localmind' is not recognized
it means Python’s Scripts folder is not in your PATH.
Step 1 — Find Scripts folder
Run:
python -m site --user-base
Example output:
C:\Users\yourname\AppData\Roaming\Python\Python312
Your Scripts folder will be:
C:\Users\yourname\AppData\Roaming\Python\Python312\Scripts
For Windows Store Python it may be:
C:\Users\yourname\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.12_qbz5n2kfra8p0\LocalCache\local-packages\Python312\Scripts
Step 2 — Add to PATH
- Open Environment Variables
- Edit User Path
- Add the Scripts folder
- Restart PowerShell
Now the command will work:
localmind init
🚀 First Run
Initialize LocalMind:
localmind init
This command will:
- Create LocalMind storage directories
- Prepare the vector database
- Check for Ollama
- Download the recommended model (
phi3:mini) - Prepare the embedding model
➕ Add Personal Knowledge
You can directly store personal information.
Example:
localmind add "My Aadhaar number is 1234 5678 9123"
Example:
localmind add "My passport expires in 2029"
LocalMind will index this information so it can be retrieved later.
📄 Upload a File
You can upload a document.
localmind upload notes.txt
The file will be:
- Parsed
- Cleaned
- Split into chunks
- Embedded
- Indexed for search
📂 Scan an Entire Folder
Index multiple documents at once.
localmind scan ./documents
LocalMind will recursively scan the folder and index supported files.
❓ Ask Questions
After storing knowledge or uploading documents, you can ask questions.
Example:
localmind ask "What is my Aadhaar number?"
LocalMind will:
- Search the knowledge base
- Retrieve relevant information
- Rerank results
- Generate an answer
Example output:
1234 5678 9123
Sources used will also be displayed.
📊 Check System Status
localmind status
Shows:
• number of stored items • index size • LocalMind storage location
🩺 Diagnose Environment
If something is not working:
localmind doctor
This checks:
• Ollama installation • embedding model readiness • system environment
📄 Supported File Types
LocalMind supports:
.txt
.md
.csv
.log
.pdf
.docx
.pptx
⭐ Best Results
LocalMind works best with clean text documents, such as:
• .txt
• .md
• .docx
Documents containing many images or scanned pages may produce weaker results, because the system primarily processes textual content.
🔒 Privacy
All data stays local on your device.
LocalMind does not send your documents or personal information to external servers.
📜 License
MIT License
👨💻 Author
Ajay Ray Samala
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file localmind-0.1.0.tar.gz.
File metadata
- Download URL: localmind-0.1.0.tar.gz
- Upload date:
- Size: 21.3 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.10
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
21fb3afd9d72e1ac983a8a69bee13005f3521c558d628e80e052312f86fb0d80
|
|
| MD5 |
d1c69475956162834f551c1461cec292
|
|
| BLAKE2b-256 |
93a23340af1cba8ef41746849a6df5934c7600c7d99d919e7350683dfe8185ac
|
File details
Details for the file localmind-0.1.0-py3-none-any.whl.
File metadata
- Download URL: localmind-0.1.0-py3-none-any.whl
- Upload date:
- Size: 24.4 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.12.10
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
00694690202ec5931b15cfa758fec12b86a381e2443aa7d6317efe8093069dfd
|
|
| MD5 |
fe46264d303c6e41c2a01ca9cbc062ba
|
|
| BLAKE2b-256 |
e65e553df430b51225518021c7f6469a7a6cb9160e804835a5e73c4687f19ee8
|