Wispr Flow's updated terms reveal a meeting notetaker in development

Why it matters

Meeting memory would give Wispr richer workplace context for its planned AI assistant, while moving the product into multi-party data collection and a heavily funded market.

Tanay Kothari (@tankots) and Sahaj Garg (@SahajGarg6) appear to be preparing Wispr Flow to record and analyze meetings, expanding the founders' voice dictation product into one of the most contested categories in workplace AI. TechCrunch reported on August 3rd that meeting features had surfaced in terms and privacy documents Wispr updated on July 25th.

Kothari and Garg started Wispr around a much harder interface problem: letting people enter text silently by mouthing words. Wispr eventually focused on voice software, releasing Flow for Mac in October 2024 before adding Windows and mobile apps. Kothari previously founded the e-commerce personalization business FeatherX, which was acquired by Cerebra Technologies. Garg led the AI team at photonic computing startup Luminous Computing after conducting generative-model research at the Stanford AI Lab, according to Wispr's founder biographies.

Wispr's updated terms define Notetaker as a set of AI features for meeting transcription, note-taking, summarization, speaker identification, action-item generation and meeting insights. The terms say inputs can include meeting audio, participant information, metadata, speaker labels and transcripts. Outputs can include summaries, action items, speaker attribution and follow-up material.

The data involved reaches well beyond a single user's dictated message. Wispr's privacy policy says the product can process participants' names, email addresses, meeting communications and calendar metadata. Wispr may temporarily retain encrypted meeting audio on a user's device or in cloud storage to resume sessions, improve transcription, check accuracy and troubleshoot the service.

Notetaker currently requires Cloud Sync, according to the terms. Wispr treats that setting separately from Privacy Mode, which controls whether customer content can be used to evaluate, train or improve Wispr's models. The terms place responsibility on users to provide notices and secure any consent required before recording or transcribing a conversation.

From voice input to workplace memory

The meeting product extends the strategy Kothari described while raising Wispr's $30 million Series A in June 2025. Wispr began with dictation, but Kothari said the longer-term plan was an assistant that understood personal context and handled tasks such as messages, notes and reminders. The original hardware work had already pushed Wispr toward capturing intent rather than producing a literal transcript.

Meeting data gives Wispr another source of that context. Flow currently turns speech into formatted text inside email, documents, messaging tools, customer relationship software and coding environments. A meeting product can capture decisions, commitments and names before a user opens those applications, giving Wispr material for the action layer Kothari has been describing.

That expansion also changes the privacy burden. Dictation generally begins when one user intentionally speaks to Flow. Meeting software processes multi-party conversations and information about people who may not use Wispr. Requiring Cloud Sync places those conversations within Wispr's cloud-based product architecture even when users disable model training.

Wispr enters a funded and crowded category

Wispr would compete with Granola, Fireflies, Read AI, Otter and Fathom. Several of those products are already trying to move beyond meeting summaries into search, email, scheduling and enterprise knowledge, the same work-assistant layer Wispr is approaching from dictation.

Wispr has raised $81 million. Menlo Ventures led a $30 million Series A in June 2025, followed by a $25 million investment led by Notable Capital with participation from Flight Fund in November 2025. That round valued Wispr at about $700 million.

A May 12th Bloomberg report republished by The Economic Times said Wispr was discussing roughly $260 million in new financing at a potential $2 billion valuation. The reported terms were still under negotiation.

A meeting notetaker would give Kothari and Garg a larger enterprise product and a reason for users to keep Wispr running through another part of the workday. It would also test whether Flow's foothold as a cross-application voice interface can provide distribution against meeting specialists that have spent years building recording workflows and organizational archives.