Another unicorn has emerged, with a maximum valuation exceeding 10 billion. How has the AI note-taking market become the most promising sector in Silicon Valley?

A new unicorn has emerged in the AI field.

Not long ago, the AI meeting assistant Fireflies.ai announced that its valuation had exceeded $1 billion, officially entering the unicorn club.

In the past 18 months, its number of active users has skyrocketed eightfold, covering 500,000 organizations and over 20 million users globally. Moreover, it has been profitable since 2023 - a rare feat in the anxiety - ridden environment of "difficult monetization of AI applications".

However, the success of Fireflies is not an isolated case.

Driven by the wave of generative AI, a group of "AI meeting note - taking" products are rapidly rising. They are no longer satisfied with just "transcription + summarization" but are evolving into intelligent hubs that can understand semantics, identify action items, connect to work systems, and drive execution.

They are redefining the concept of "recording" and attracting close attention from investors:

Granola: Starting from meetings, it aims to be the "second brain for personal office work", with a valuation of $250 million.

Notion: It builds a Lego - style work platform, with a valuation of over $10 billion.

Abridge: Focusing on the medical system and reconstructing doctors' notes, it has a valuation as high as $2.75 billion.

These three companies represent three evolutionary paths of AI note - taking products and reveal a common signal:

When AI becomes the "cognitive infrastructure" of modern organizations, recording, as the starting point of information input, is shedding its past tool - like nature and becoming the entrance to the collaboration system and a key node of organizational intelligence.

01 Fireflies AI: Not Just a Meeting Recorder, but an All - Around Team Assistant

As a leading player in the field of AI meeting recording tools, Fireflies AI has solid "basic skills" that inspire confidence: it supports 69 languages, with a real - time transcription accuracy of up to 95%. It can complete meeting transcription, summarization, and action item extraction within minutes.

However, this is just its "starting point".

What truly sets Fireflies apart from its competitors is its powerful ecological connection ability and AI collaboration engine. It doesn't just take notes; it is reconstructing the entire team's information flow and execution flow.

Fireflies can seamlessly integrate with more than 40 office tools such as Zoom, Google Meet, Teams, Salesforce, Slack, and Notion. It automatically extracts key action items, customer requirements, and technical decisions from meetings and synchronizes them to various project management systems.

In a nutshell, it bridges the gap between "understanding" and "doing", making meetings not just about recording but directly driving execution.

For example:

Customer requirements mentioned in a meeting automatically generate tasks and are written into Slack to remind relevant responsible persons.

Technical solutions discussed by the team are automatically written into Notion for memo.

Details of sales follow - up are entered into the CRM system with one click.

Meanwhile, Fireflies has also connected to Perplexity's search engine, enabling real - time internet access. Users can ask questions during meetings, such as "Fireflies, what is the current market forecast for AI meeting assistants in the industry?" The system generates answers instantly, truly bringing AI into the discussion.

Furthermore, Fireflies plans to launch a "Talk to Fireflies" dialogue agent and develop an AI proxy that can "attend meetings on your behalf". It can even retrieve context from your historical meetings and actively supplement missing information.

Currently, Fireflies has launched over 200 "AI agents" for different functions, suitable for different roles and application scenarios, such as sales, marketing, recruitment, operations and management, and customer support.

In the recruitment process, after an interview, the recruitment team can use Fireflies AI applications (such as the "cultural fit extractor" and the "candidate feedback aggregator") to generate candidate scorecards.

In the marketing process, the product launch planning application can formulate product launch strategies, and the event performance evaluation application can extract key insights from event discussions.

In the sales process, the sales application can extract detailed information such as the sales team's budget and schedule.

In the operations process, Fireflies AI applications (such as Standup Notes) can capture action items and obstacles, and the goal progress tracker can monitor progress.

In the after - sales process, the agent performance feedback generator can generate insights into customer service agent performance and provide guidance suggestions.

Its underlying logic is clear: each department has different languages and collaboration methods, and Fireflies understands these "departmental dialects" through customized agents and helps you automatically execute tasks after meetings.

In the current situation where AI meeting notes are becoming increasingly "tool - centric and competitive", Fireflies has clearly crossed the boundary of recording and is striving to become a real intelligent collaboration hub that can "understand human language and drive the team".

02 From Tool to Hub: The Breakthrough Path of AI Note - Taking

A scribe is an ancient profession dedicated to recording things for others.

In ancient times, this profession also took on the roles of secretary and administrator, such as recording orders for the royal family and temples and managing daily affairs. Later, scribes also worked in fields that required a lot of language communication, such as business and law.

Today, the need for recording is ubiquitous in work and life, for example, in the medical, legal, and government sectors. As shown in the figure below, in the US transcription market, there are approximately 100,000 scribes in the medical field alone, accounting for 1/10 of the total number of doctors. Research shows that the market size of the US transcription market reached $21.6 billion in 2020 and increased to $2.38 billion in 2021.

With the rise of AI, this ancient industry, traditionally driven by human labor, is being completely disrupted. The author noticed that in addition to Fireflies.ai, a group of AI meeting note - taking startups have emerged overseas. They have not only achieved remarkable growth but also received a lot of financing.

So, how do these AI meeting note - taking companies bypass the applications of large companies and find their own differentiated paths in the fierce competition?

Granola, the "Second Brain" Valued at $250 Million

Granola is one of the most notable rising stars in this field.

Judging from the data, the number of users of this product grows by 10% every week, and the monthly retention rate is as high as 70%. On May 14th, Granola announced the completion of a $43 million Series B financing, with a valuation of $250 million.

Granola chose the high - frequency scenario of meetings as its entry point and made a lot of innovations in product and interaction.

Granola positions itself as the user's "second brain".

Different from most AI note - taking products that directly generate relevant meeting minutes, Granola chooses to co - create with users. That is, users can record any notes or fleeting thoughts at any time, and Granola will capture the conversation content in real - time and transcribe it. After the meeting, it will further improve and enrich the transcribed content based on your notes.

In essence, Granola divides the work between users and AI very well. Humans record what they think in their minds, and AI records what is said in the meeting. Finally, the large model combines the two.

This approach not only fits the real - life meeting scenario better but also resonates more with users. This design precisely meets the "hybrid thinking" needs in real meetings and gradually transforms Granola from a note - taking tool to a lightweight collaboration platform.

Notion, the Lego - Style AI Note - Taking Platform Valued at Over $10 Billion

Compared with Granola's approach of entering the market through the meeting scenario, Notion chose a more difficult path - to build a "general work platform".

Notion's positioning is to piece together the functions you want with "Lego bricks". Documents, tasks, databases, and even AI assistants are all modular blocks. Users can build their own workflows like building Lego.

Specifically, Notion's features are reflected in the following aspects:

① Free switching: The platform integrates a large number of AI functions, so users don't need to switch back and forth between the platform and AI tools.

② Arbitrary text dragging and dropping, modular functions: Notion allows users to paste content between different editor tools. It can turn functions such as documents, databases, and tables into Lego - like modules, allowing users to build whatever they want flexibly and with strong expandability.

This reflects Notion's Lego - style product philosophy, which uses a block - based rather than an application - based approach to build its software system.

However, the other side of taking the platform route is that without a natural industry moat, it must win space through user perception, product freedom, and collaboration depth.

At least for now, Notion is doing well. With its powerful AI functions and convenient flexibility, Notion has increased its user base from 1 million to 100 million in just over three years, and its valuation exceeds $10 billion. Its investors include top - tier institutions such as First Round, Index, and Sequoia.

Abridge: Deeply Rooted in the Medical Field, an AI Recording Assistant for Doctors

Abridge chose a more professional but highly potential direction: the medical scenario.

Different from lightweight products targeting individual doctors, Abridge directly cooperates with the US medical system and becomes part of the electronic medical record (EMR) system. It has entered more than 100 authoritative institutions such as the Mayo Clinic, Duke University, and Yale, covering tens of thousands of clinical doctors.

Its technical foundation is also fully customized for the medical field - a self - developed large - scale speech recognition model, trained and fine - tuned for dozens of languages and hundreds of professional fields to ensure that the note content is accurate in clinical terms, well - structured, and in line with doctors' writing norms.

Supported by the proprietary model, Abridge supports accurate clinical summarization and medical terminology in more than 50 professional fields and can identify, understand, and accurately record notes in more than 14 languages, ensuring the integrity and high precision of structured clinical note drafts.

In February this year, the AI medical note - taking company Abridge completed a $250 million financing, with a latest valuation of up to $2.75 billion, led by well - known Silicon Valley venture capitalist Elad Gil. Considering that Abridge's ARR was only $50 million at the end of 2024, its EV/ARR is as high as 55 times, even higher than that of large - model companies.

AI note - taking products that deeply integrate professional processes, such as Suki AI, Nabla, and Freed AI, are reconstructing the generation logic of traditional medical documents.

03 Conclusion

Granola, Notion, and Abridge represent three evolutionary paths of AI note - taking products:

Starting from meetings and becoming the team's second brain: like Granola, emphasizing human - AI co - creation.

Building a platform and becoming a general work hub: like Notion, emphasizing Lego - style combination and ecological integration.

Deeply penetrating into a vertical industry and reconstructing professional scenarios: like Abridge, emphasizing data structuring and system integration.

Their commonalities are:

1) They all move from "recording" to "understanding". They are not satisfied with simple voice transcription or meeting minutes but use AI to understand user language, extract structured content, and identify behavioral intentions, upgrading from information recording tools to intelligent knowledge assistants.

2) They are all building "bigger containers". They don't create "isolated tools" but seamlessly integrate with users' existing systems through plugins, APIs, or native blocks and integrate into the core organizational processes.

3) They are all based on the product philosophy of "role understanding". All three companies realize that different user roles have different languages, tasks, and cognitive methods, so they all emphasize "design for roles".

In short, the common direction of these AI note - taking products is to make AI go beyond the recording level and penetrate into the "neural network" of organizations, becoming the infrastructure for decision - making and collaboration.

This article is from the WeChat official account "Wuya Intelligence Talk", author: Intelligent Crow, published by 36Kr with authorization.