Fellow

Fellow is an AI-powered meeting assistant designed to streamline the entire meeting lifecycle—from preparation and real-time documentation to post-meeting follow-ups—while placing a strong emphasis on data privacy and security. Unlike generic transcription tools, Fellow integrates directly into popular calendar systems, video conferencing platforms, and project management apps, allowing teams to automate note-taking without switching contexts. The core functionality revolves around three phases. Before a meeting, Fellow syncs with your calendar to pull agenda items, attendee lists, and relevant documents. It can pre-populate shared meeting notes with talking points, previous action items, and context from past conversations, ensuring everyone arrives prepared. Users can also set privacy controls at this stage, deciding who can view, edit, or access the meeting notes and recordings. During the meeting, Fellow acts as a silent observer. It records audio, transcribes speech in real time with high accuracy (claimed to be near human-level), and identifies speakers. The transcription is not simply a text dump; Fellow automatically distinguishes between casual discussion, decisions, and concrete tasks. It can highlight questions, votes, and key moments. Importantly, the AI processes everything locally or within the user’s controlled environment—Fellow states that it never trains its models on customers’ meeting data, and all recordings and transcripts are encrypted both in transit and at rest. Users have granular permissions: a meeting host can restrict access to the recording entirely, limit viewing to specific participants, or allow external guests temporary access. After the meeting, Fellow generates a structured summary that includes: a concise overview of the discussion,