Vellum AI

Vellum is a personal AI assistant that now runs natively on iOS and Android, extending an open-source platform that previously lived on macOS, the web, Slack, Telegram, and other surfaces. Vellum cloud hosted enables assistant memory to persist across every surface, so context never resets when you switch devices, which is the point of shipping mobile apps: the same assistant that handled work at your desk can update product listings, pull payroll hours, or answer a quick question from the sidewalk. The launch also carries over the features from desktop, including the ability to run speech through the same agent loop as text chat, with interrupt handling, plus a choice of local or cloud model backends. The product is aimed at people who want one assistant that follows them everywhere and stays under their control. The self-hosted version is open source and can be deployed on local hardware, a VPS, or private cloud , which sets it apart from closed mobile assistants tied to a single vendor. Vellum is an open-source personal AI assistant that can run persistently across Mac, iOS, web app, voice, email, Telegram, and Slack , and the mobile release is what closes the loop for founders, operators, and anyone whose day happens away from a keyboard. Vellum AI is built by a team led by CEO Akash Sharma with Co-CTOs Noa Flaherty and Sidd Seethepalli , alongside founding engineer Vargas Fuertes. The mobile launch matters now because the assistant category is consolidating around chat interfaces locked to specific labs, and Vellum is betting that a portable, model-agnostic, open-source assistant with persistent memory across every device is the alternative worth carrying in your pocket.