Top 6 agentic features in meeting assistants

Meeting assistants, traditionally focused on recording, are now advancing to more dynamic roles with agentic features that take actions on behalf of users during or after meetings. This shift is driven by improvements in real-time multilingual transcription and large language models (LLMs) that enable accurate and timely interventions. New functionalities include live in-call lookups, real-time intervention, autonomous follow-up, stateful reasoning across meeting histories, meeting context as a service to other AI tools, and source-grounded synthesis. These features are transforming meeting assistants from passive recorders to proactive participants that can retrieve information, alert users about missing data, draft follow-up communications, and synthesize information into consumable formats. Companies like Fireflies, Abridge, Gong, Sana, Granola, and NotebookLM are at the forefront of these developments, leveraging their respective technologies to enhance productivity and decision-making by integrating meeting data seamlessly into broader workflows. The evolution of these tools is underpinned by the quality of speech-to-text models, which form the foundation for all agentic features.