For a product that promises to kill the meeting, the business of meeting automation is surprisingly crowded. The pitch is familiar: an AI bot joins your calendar, records, transcribes, and summarizes the call, then pushes a digest to your project tools. What’s less clear is how any one player builds a lasting wedge in a market where the core transcription service is fast becoming a commodity. MeetGeek, a Bucharest-based startup, is making its bet by moving the AI from the sidelines into the conversation.
Founded in 2020 by three technical co-founders, including CEO Dan Huru, MeetGeek has grown on the standard playbook [Crunchbase]. It automates notes across Zoom, Google Meet, and Microsoft Teams, supports over 50 languages, and integrates with Slack and Notion [MeetGeek.ai]. Public estimates suggest the company reached an estimated $1.5M to $2M in annual revenue with a team of around 12 [getlatka.com] [businessforum.ro]. In September 2025, it secured a €1.6 million seed round led by Early Game Ventures to scale its AI workspace [ain.ua]. The reported traction is a tripling of ARR over the past twelve months [businessforum.ro].
From Automated Scribe to Active Participant
The company’s answer is a shift from passive recording to active participation. In April 2025, MeetGeek introduced "Notetaker Avatars," a visual refresh framed as a first step toward more interactive, agentic AI [support.meetgeek.ai, 2025-04]. By October, it had launched AI Voice Agents capable of autonomously joining virtual meetings to speak, ask questions, and even lead discussions [cbinsights.com / manilatimes.net, 2025-10/11]. This is the strategic pivot. Instead of just documenting what happened, the AI is being tasked with shaping what happens next.
The ICP and the Competitive Set
MeetGeek’s ideal customer profile is the distributed, client-facing professional services team. Think consultancies, agencies, or product research firms where meetings are the primary unit of work and billable time. The mobile app for recording in-person conversations extends this use case to offline coaching and interviews [MeetGeek.ai]. The platform’s analytics on participant engagement and sentiment are aimed at managers who need to improve team dynamics [MeetGeek.ai].
Its realistic competitive set breaks down into two tiers. The first is the direct feature competitors like Fireflies.ai and Read AI. The second, more formidable tier is the platform giants. Microsoft, Google, and Zoom are all baking similar AI note-taking and recap features directly into their core collaboration products. For MeetGeek, differentiation must come from deeper workflow automation, superior analysis, or a unique agentic layer that the platforms are slower to ship.
The Execution Hurdles Ahead
The bet on voice agents is ambitious, but it introduces several new execution risks:
- Adoption friction. Asking a team to trust an AI voice to speak in meetings is a bigger behavioral leap than letting a silent bot take notes.
- Commoditization pressure. The foundational transcription layer is a race to the bottom. MeetGeek’s pricing starts at $9.99 per month [MeetGeek.ai], but competing on price against bundled platform offerings is a losing game.
- Scale and support. Tripling ARR is a strong signal, but doing so from a small revenue base with a team estimated at 12 people [getlatka.com] means the company is stretching thin.
The Next Twelve Months
The coming year will be about proving that the voice agent is more than a feature demo. Success will be measured by whether MeetGeek can convert its existing base of note-taking customers into paying users of its agentic layer, and whether it can land its first enterprise deals where the AI is contractually tasked with facilitating specific meeting types.