Sanctum is a downloadable Mac application that enables users to run full-featured open-source large language models (LLMs) locally, ensuring that all data processing occurs on-device without requiring an internet connection. This approach addresses privacy concerns by encrypting user data and storing it locally, preventing any external access or data leakage.
Funding
Funding not disclosed
Founders
Product
Problem
Cloud-based large language models (LLMs) raise privacy concerns due to the potential for data breaches and lack of user control over personal information. Businesses and individuals handling sensitive data require a secure way to leverage AI without exposing their information to third parties.
Solution
Sanctum is a desktop application that enables users to run open-source LLMs locally, ensuring data privacy and security. The application encrypts user data and stores it directly on the device, preventing external access and data leakage. Users can interact with AI, analyze documents, and perform data analysis without an internet connection, maintaining complete control over their information. Sanctum integrates with Hugging Face, allowing users to easily access and switch between thousands of GGUF models.
Target Audience
Sanctum is designed for professionals such as legal experts, financial analysts, academic researchers, and medical personnel who handle private data and require secure AI tools.
Features
- Local processing of LLMs directly on the user's device
- On-device encryption using AES-256 to protect data within a secure vault
- Compatibility with MacOS 12+ and Windows 10+
- Hugging Face integration for seamless access to GGUF models
- Ability to chat with PDFs in a secure and private environment
- Offline functionality, allowing operations to continue without an internet connection