If you are often taking interviews or simply taking notes on the go, there's a good chance you rely on a transcript app or keep on the lookout for such an app.
With modern AI breaking through in the last two to three years, note-taking and voice transcription have become major use cases. Fireflies.ai has been one of the names associated with this. The company recently sponsored T20 league ETPL and also announced Talk, which lets users dictate across their Email, Slack, documents, CRM fields, and other applications on Windows as well as macOS. We had a quick chat with the company co-founder & CEO, Krish Ramineni:
Other than Email Assistant, what are some avenues the company is exploring to bridge the gap between AI note-taking and people who have never used such an app?
We're expanding well beyond note-taking. The first area is voice agents. For years, Fireflies would join your meeting and take notes for you. We're now building towards Fireflies attending certain meetings on your behalf: gathering information, answering questions, and bringing back the things that matter to you.
The second is Email Assistant, where Fireflies understands your inbox, helps draft responses, follows up, and takes the load off managing everyday communication. The third is Fireflies Talk. Meetings aren't the only place people lose time to typing. With Talk, you can speak into Fireflies almost anywhere and get clean, structured text in your email, Slack, documents, or ChatGPT. We think voice will replace a large share of keyboard use over the next few years.
The bigger goal behind all of it is simple. Nobody should need 50 different AI tools for meetings, email, dictation, scheduling, search, and the repetitive work in between. We want to bring all of that into one AI teammate that already has your context.
What are the biggest challenges for somebody like Fireflies.ai since modern AI took a huge step during the past three years that they may have not seen before?
The biggest shift is that things which used to be technically very hard have become easy. Three years ago, transcription, summarisation, and understanding conversations at scale were real technical advantages. The underlying models have improved so fast that those capabilities are now within reach of almost anyone. The bar has moved. A good model is table stakes today. What's hard is everything around it: distribution, integrations, reliability, security, workflows, and a product people actually want to open every day.
Trust has become just as important. We handle some of the most sensitive information inside a company: meetings, customer conversations, email, and internal knowledge. Customers want to know where that data sits, how it's stored, whether it's being used to train models, and what controls they have over it. We've invested heavily on that side. We don't train on customer data, and we've built for enterprise-grade security, privacy, and admin controls. So the challenge isn't keeping up with the next model release. It's taking all of this rapidly improving AI and turning it into something people trust enough to use every single day.
Can you give us a breakdown of your current user base in terms of operating systems or platforms in use?
We don't really look at our user base by operating system. We look at where the work is happening. Fireflies runs across four surfaces: the Web product, a desktop app that captures meetings directly from your computer, a mobile app for in-person conversations, and a Chrome extension. On top of that, we have more than 120 integrations across video conferencing, CRM, and collaboration tools.
In practice, the same person might use Fireflies in a Zoom call in the morning, switch to the mobile app for an in-person client meeting in the afternoon, and have both sets of notes pushed into Salesforce automatically. The operating system barely matters to that journey. That's deliberate. Fireflies should work wherever the conversation is happening, rather than forcing someone into one device, one OS, or one workflow.
How does Fireflies.ai see its user base increase with investment into various sports tournaments?
We don't treat sports as a user acquisition channel, and we don't measure it that way. It's about brand awareness, and about reaching people who sit outside the usual SaaS and tech audience. Fireflies already has more than 20 million users, but there are still millions of people who have never used an AI note-taker, or who haven't stopped to think about what an AI teammate could actually do for them at work.
Sport is one of the few places where you can reach that many people at once, and where the idea translates without any explanation. The ETPL is a good example: our branding ran right through the Rotterdam Dockers season in front of cricket audiences across Europe and India. So, I wouldn't judge a tournament by how many sign-ups it generated that month. The value is that millions of people now recognise the brand and understand the idea, so that when they do need an AI teammate at work, Fireflies is the name they think of.
