Vellum
Personal AI assistant with long-term memory that runs locally or in the cloud and takes real actions across your tools on a schedule.
Visit website
What Vellum does
Vellum is a personal AI assistant built around long-term memory. It remembers you, your preferences, the people in your life, and your work across every conversation, so you never have to repeat context. By default it runs locally on your Mac, keeping your data on your device and off Vellum's servers unless you opt in. Beyond answering questions, Vellum takes real actions: sending emails, managing your calendar, browsing the web, running background tasks, and more. It works across Mac, iOS, Android, web, Slack, Telegram, and voice, and can be extended with community plugins or custom skills.
Vellum features
- Long-term memory built to remember you, your preferences, and work context across all conversations
- Runs locally by default with your data never used to train models unless you opt in
- Autonomous actions across email, calendar, Slack, Notion, files, and the web with a fine-grained permissions system
- Available on Mac, iOS, Android, web, Telegram, Slack, voice, and terminal
Who Vellum is for
- People who want a private, locally running AI assistant with real long-term memory
- Users who want scheduled background automation across Google Workspace, Slack, and Notion
Vellum FAQs
Can Vellum take actions without my permission?
Not by default. High-stakes actions like sending email to new contacts or modifying important files require explicit approval until you mark them as trusted. You stay in control throughout.
Where does my data go?
By default, your conversation history, memory, and credentials stay on your machine and are never used to train models. Vellum is open-source so you can inspect the code. If you use Vellum Cloud, your assistant runs in an isolated container provisioned exclusively for you.
