Interview with Fireflies.ai
CEO and Founder Krish Ramineni.
On the latest edition of the Silicon Valley Momentum Podcast, Roland Siebelink of midstage.org spoke with
Krish Ramineni, the cofounder and CEO of Fireflies, an AI voice assistant. Fireflies has similar
functionality to systems like Siri or Alexa but is designed for the workplace. During the podcast, you will
hear from Krish about how Fireflies is the perfect workplace assistant while so many people are working
remotely amidst the COVID-19 pandemic.
You will also hear the whole conversation between Roland and Krish that covers topics like:
- How a voice assistant wasn’t the initial priority when Krish and his cofounder first created Fireflies.
- Why Fireflies is yet to hire any workers in the sales department.
- The key distinctions between building a product and building a team when things start to take off and
accelerate faster than expected.
- The importance of use cases as a way to learn about the market.
- How Fireflies has achieved organizational virality.
Roland Siebelink: Hello, everyone, and welcome to the Silicon Valley
Momentum Podcast. I'm very lucky today to have as our special guest Krish
Raminemi, who is the co-founder and CEO at Fireflies.ai, one of the most
compelling voice assistants out there.
Hello, Krish, thank you for joining.
Krish Ramineni: Hi, Roland. Great to be on the show and thank you so much
for taking the time to have me here.
Roland Siebelink: I'm so excited to have you here because in these pandemic
times that we're currently living through, we've seen some amazing traction with
voice assistance and everything that supports remote work. I'm looking to get
into that very soon. Krish, when you would describe Fireflies in just one quick
elevator pitch, what problem do you solve for which target group and what's your
Krish Ramineni: Fireflies is a meeting assistant that joins your meetings
across various video-conferencing platforms. It's able to capture, transcribe,
and help you essentially complete work in systems that you are already in.
Whether that's a CRM like Salesforce, a chat system like Slack, or a project
management system of your choice.
Our whole goal has been to be able to help people focus on their conversations
and let AI complete work in the background and perform tasks for them.
Roland Siebelink: You're not the only voice assistant in the market at this
stage, right? What differentiation do you have compared to some of the other
companies in this field? Those that shall not be named, of course. Do you target
a different group of customers? Do you have a specific feature that others don't
possess? What's your differentiation?
Krish Ramineni: I think the market is very uniquely segmented. On the
consumer end, you have the Siri, Amazon Alexa's of the world, Google home, which
don't directly compete with what we're doing. They are really at-home
assistance. What we are essentially doing is bringing some of the same
capabilities and functionality to the workplace. Specifically, your meetings and
the work you're doing.
The use case is very much on the B2B work side. The other areas of voice, in
general, have been a very interesting space. A lot of the core technologies
people are investing in ASR, the speech recognition transcription, I think
there's going to be commonalities on how technology is built.
A decade ago, everyone was building on mobile. Or, 20 years ago, everyone was
building on cloud. But I think one of the key differentiators in terms of what
we focus on is helping with workflows and allowing people to complete tasks
versus primarily focusing on any single vertical and doing analytics.
I think there are tools out there that do a great job of performing sales
analytics or customer support analytics. Whereas for us, we take more of a
direction of how do we help people complete work and tasks across different
verticals. Whether you're in sales, whether you're a product manager, whether
you're an engineer or someone in ops, it's more of a work assistant.
Roland Siebelink: Okay. Would that mean that you're trying to sell to an
entire company at large or just any employee that finds you? What's your go-to
Krish Ramineni: For us, we focus on the end user and the individual user
needs to have an incredible experience. For us, if you go to our website, our
pricing is transparent. People can sign up and start using Fireflies
immediately. And they can bring it to their teammates and their colleagues and
start inviting their teams as well to start using it.
And from there, Fireflies proves its utility. There's different use cases people
explore with it. And then over time, the goal is you start seeing Fireflies
spread end to end within an organization. You might see cross functional teams
using it. You might see small organized, small departments using the platform.
There really isn't, I think, a restriction in terms of how it spreads. It truly
is a horizontal platform with numerous use cases, very similar to the slacks of
the world or the zooms of the world.
Roland Siebelink: Awesome. I definitely want to get back to that a little
bit when we talk about traction. But maybe also for context, can you tell us a
little bit about how Fireflies came into being? I believe you were founded at
MIT. Is that right?
Krish Ramineni: Yeah. Both my cofounder and I had started Fireflies in
Boston. Before that, I was a product manager at Microsoft, working on the
collaboration suite and it was just a wonderful experience. Afterwards, I was
actually planning on going to grad school and I had a little bit of time before
going to grad school. And so I went back to meet one of my friends, who I had
met several years ago while I was in college.
He was finishing up at MIT, Sam. Sam, my co-founder and CTO. He was doing
machine learning, deep learning. A lot of interesting things in that space at
MIT. And we got started there. We spent some time in MIT, in Boston, to get the
company rolling. And then we moved out to San Francisco, where I'm originally
from, to set up our HQ.
Roland Siebelink: You already mentioned the technology investment in ASR and
other voice technologies. What other trends and changes are you seeing that
you're capitalizing on? And how would you describe your future market and your
Krish Ramineni: Definitely when we were starting out, voice wasn't our first
priority. We were thinking of building more chatbots and email bots that can
help understand conversations. A few years ago, there was a huge trend around
bots. And we saw the rollout of the Slack bot ecosystem and you see bots on
websites when you go and interact with the website on their live chat. That was
definitely our area of interest and building workflows around that.
What we started to see was there was this saturation in the chat bot and email
bot and workflow tools. But voice was a blue ocean. A lot of people weren't
tackling it. But we also felt 70% of all the major conversations that ever
happened in an organization happen inside meetings and voice conversation. That
felt like a much bigger market, a much bigger blue ocean for us to adapt and
dive into. It's definitely a much harder market because the underlying
infrastructure and technology needed to do well in that market takes a lot of
effort and time.
But we felt that could be a great starting point to helping organizations unlock
all these conversations and all these important insights and knowledge from
this. That's why we decided to go the route of voice. And as a byproduct of
that, the technology, the voiceover IP, the infrastructure, had to be built in
order to be able to scale to that.
Our initial hypothesis always was that there is gold buried inside
conversations. There is knowledge buried inside conversation. We stuck to that
theme, but there were many iterations and pivots throughout the process. I think
the thing that you start seeing is before we even started building the product,
we knew something like this should exist.
There are large companies that wanted to make these things exist. Whether you
look at the Googles of the world, the Microsofts of the world, I think there's
always this vision of, "Oh, it would be incredible to have a meeting assistant
that joins as a participant and does work for you like an AI secretary."
That theme has been going on for, I would say decades. The interesting thing is
how do you actually make it work? How do you actually market with something that
provides value? And I would say the technology has also been trending in the
right direction to enable this, right? Voice technology has come a long way in
the last five years. Some of the work we do today, maybe five years ago, would
have required a dozen PhDs to really dive deep into.
Now you have the democratization of AI. You have better models. You have better
GPUs and tools to run these systems on. And so now you're starting to see the
commercialization of this technology. And so for us, it's also the technology
wave that we're able to ride. It's definitely something even two to three years
ago, the technology risk was very high. And it's not completely gone, but it's
something that has dramatically improved, especially over the last two years.
Roland Siebelink: Okay. Is that also what gave you the courage, you and Sam,
to go into this market knowing full well, as you said, that behemoths like
Microsoft and Google? Microsoft you knew from the inside, of course, that they
were also looking at that market as something that should exist.
Krish Ramineni: I think from the likes of Microsoft and Google, I myself
wasn't working on anything directly related to this stuff at Microsoft. But
there was always this fascination when you hear talk from folks about what does
the future of work look like? And they talk about augmented assistants and I
almost think every company has some sort of narrative related to this. It's just
really cool to be able to go out and take that sort of grandiose vision, distill
it down to the parts that matter, and then being able to build something out
that people actually want to use, right?
I think what I've learned through this whole process is there's things people do
on the marketing front and the narrative there's that people sell that convinces
folks. And then at the end of the day, great ideas just need constant
distilling. You get down to the core parts and that's where, and then you go
through that distillation and then you build the product and then you start
realizing, okay. "Oh, there's a lot more that is now possible with this now that
it's on the market."
It's almost like an hourglass where it has to go through the distillation
process and then execute on that. And then as you start scaling, you start
realizing, "Oh, there's way more potential to this than we had initially
Roland Siebelink: Don't leave your ideas just in your head or in your deck
because you'll never actually get to learn the reality of the ideas before you
have it on all the market in a prototype version, at least. That's what you're
saying, I believe, right?
Krish Ramineni: Yeah. You don't know until the rubber hits the road and
there's customer feedback where this is even going. And that's a lot harder than
it sounds, right? Getting folks to product market fit and then beyond is
definitely a very hard stage for a lot of startups, including ourselves when we
Roland Siebelink: What were some key things you did to get through that
stage? Everyone struggles with finding product-market-fit and then getting to a
scaling stage, as you say. What were some things, in retrospect, that you did
right? And maybe there's also something you can share where you spent way too
much time or effort on this and it didn't pay off to us in the end?
Krish Ramineni: I think from day one we wanted to build a product that was
self-serve and that's something people can go use, realize value, get activated,
and then swipe their credit card and start using. For us, that was a very big
decision. We, today, don't have any sales people. That is something that might
change over time. But it was really important for us that, in order for this
technology to succeed at the scale that we envisioned, we needed people to be
able to start using it and realizing value from it.
We just hyper-focused on getting people to start using the product, discovering
it, learning things. And then there's no silver bullet. You just have to keep
chipping away, keep iterating away at all the bottlenecks. From the customer
acquisition to the activation funnel to the recurring usage funnel.
And then that's the less glamorous part. But that is also the most important
part that you have to keep setting up and keep iterating on. I think it's very
different versus if you have a sales-driven approach where you can sell the
vision, you can have someone evangelize it, you can do solution selling, you can
do a lot of interesting things in that regard and then go to market that way.
There's no right or wrong answer. Given that we were very
product-engineering-centric founders, we felt like that the best way is that
we're going to build something that people find value from. And so that's the
direction we went in. And that was, honestly for us, the best way to identify
what people care about and what people don't care about.
Roland Siebelink: Excellent. You said you have no sales people at the
moment, right? So it's a completely product-led-growth model, as I understand
it. Talk to me, a little bit, about the traction you guys have seen so far,
especially in these special times with so many remote workers.
Krish Ramineni: Yeah. In the last six months, especially the current version
of Fireflies and what it's been through - we rolled out the recent version in
early January and then Covid hit in March. I would say we were definitely COVID
beneficiaries for what has happened. There's this intersection of so many
themes, right? Voice, like we talked about, remote work, and collaboration
software. You've seen Zoom go through the roof. Google Meet, Microsoft Teams,
and all these platforms that are on the video conferencing site go through the
roof. I read somewhere that in a very short period of time, 60% or more of the
country just went into remote mode.
And all of a sudden that has changed the nature of how people think about tools
like Fireflies. And we were already part of crossing the chasm and then that got
accelerated. I know that the remote workspace itself has been growing at almost
150% in the last decade. And this was before COVID hit. It's just, I think, a
really interesting period in time that no one could've planned for.
And it's funny because when we first started the company, we didn't even use the
word remote in our products. We said, it's going to be something that helps
people at work and a meeting assistant that joins your conference rooms. And we
still have that. But it's just so fascinating when I go online and see people
talk about the space, they include Fireflies in their market landscapes. And
they say this is one of the remote work products that's become popular over the
last couple of months.
It was just great to see that the market pulled this out of us versus us having
to engineer a market around what we do. That, honestly, is a blessing in
disguise more so than anything.
It's definitely something that took us by surprise, especially starting around
February and March, as the acceleration happened. But it also was just part of
our roadmap or our initiative that we were going after. And there's a lot more
value we want to provide to businesses and teams that use Fireflies beyond what
the system offers today. But it's a great thing to see.
All we see is constantly having to keep up with servers and scale and just
making sure that the system’s up all the time and go through growing pains,
which is very common for a lot of startups. But it's only when you're in that
hair-on-fire mode do you realize that, “Oh, it's the market that is really
pulling this out of you.” You need to hire people. You need to delegate
responsibilities and you need to move from just building a product to building a
team that can handle all of this that's happening.
I think that's definitely happened at a much faster pace than we had
anticipated. And that's just a growing pains that you deal with.
Roland Siebelink: Yeah, absolutely. Can you talk to me a little bit about
your team? I know that founders are typically interested in what kind of people
do you hire? What profiles? What are you prioritizing? And also, how do you make
sure that the people you hired are actually the right people for your team?
Krish Ramineni: Yeah, that's a great question. One interesting thing about
us is we're a 100% globally distributed team across several locations in
multiple countries. And so, from day one, even though we didn't use remote work
in our pitch about Fireflies, we ourselves wanted to build a remote working team
because that's how my co-founder and I started working for a long period of
And it just felt like sometimes engineers just want to be able to focus. It was
out of convenience more than anything that we started that way. And today, you
see companies like Zapier, GitLab, adopted at a much larger scale.
And so, we are a remote team. As a result, we also are one of the largest dog
feeders of our product and are able to make improvements based on how our team
grows and evolves. The team today is primarily engineering, product, and design
focused. And a lot of folks, just thinking about voice, a lot of people are
thinking about scalability and infrastructure.
Our next goals are to hire for our go-to-market. Bringing in our first marketing
hire or bringing in our first sales hire down the road. But a lot of what we've
done has just been about refining the product and the overall experience.
I think it also helps us hire a little bit easier because a lot of engineers and
product folks want to join a company that is product-first rather than
sales-first. For them, this is music to their ears in terms of how we give them
responsibility and ownership in terms of driving things. And they are very close
to the customer and hear the voice of the customer.
For us, that DNA is there. We're very engineering and product centric. There's a
time and place for us to translate that into marketing and sales. But I'm always
a believer that sales and marketing becomes a lot easier if you're selling
something that people really want and it already has a high demand for it.
Roland Siebelink: What I would say is that I hear from other founders that
what they would worry about if they didn't have such emphasis on marketing and
sales, is the competition going to catch the market before we do? How do you
guys feel about that?
Krish Ramineni: Yeah, that's definitely a great question. I think because of
our use cases and our horizontal application, we felt this is the right way to
go about it. If we were focused on a singular vertical, like sales or customer
success or support, and building a very tailored product for those use cases
like some of the other folks in the voice market, it absolutely makes sense for
them to double down on sales. Double down on the marketing and raising lots of
money to be able to go do that.
We're in an interesting market where we're focused on a lot of these use cases,
a lot of bottoms-up adoption, and then there will be a lot of workflows and
other opportunities for people to adopt the platform on. That definitely helps
us in terms of this. And I think it's almost a blessing to have these sort of
use cases and knowledge so that we can learn from it.
I think, also, the other trade off is when you go down that particular vertical
focus, which I think is great, you tend to have to price a lot higher. And you
have to go after a user segment that's willing to pay a lot so that you can
justify the sales and marketing expenditure.
That also becomes an interesting thing where you play this pricing game. In
terms of what we're trying to do, and in terms of how you can buy Fireflies
today, an individual can buy it on their own credit card. A team can buy it on
their team's expense budget for the month. There is a very different strategy
that we're going. And we believe in the model like at Lacien and Slack where you
go for volume. You go for a lot of users and lots of seats and lots of adoption.
And when you can do that, that's when you get a lot of use cases in a user base
that you can build on top of.
Roland Siebelink: All right, very good. We also said something before about
the virality. This time, not about the Coronavirus but about the apps spreading
like wildfire throughout an organization, right? Can you talk a little bit about
that? Because so many founders would love to have the B2B software that gets
signed up by one, two individual contributors and then starts having that
organizational virality. Any tricks they can learn from? Any insights that you
Krish Ramineni: Yeah, I think there's definitely certain things that we've
seen. Sharing is a very good indicator. If people are able to share their
meetings or recaps with other teammates, that always helps light the fire.
Number two is the value that a person receives before they even sign up for
Fireflies. When someone else sees Fireflies in a meeting like this, and if I had
never used Fireflies, I'm going to naturally ask you about what is the product.
And if I'm able to experience the notes or the recaps and transcripts before I
even sign up, then I've been activated. And that is always a great, great way to
And then like with anything, there's virality you can track through people
sharing and people signing up or teammates signing up or other organizations
signing up. And then there is something to be definitely said about virality
through word of mouth. When people tweet, when people talk about the product,
that, I think, is the ultimate sign. It doesn't matter if you're a bottoms-up
SaaS business or you're a traditional SaaS business where you have free trials
and you need to get people through.
Those are definitely really important pieces of the funnel where you need people
to be able to be your advocates. And we take every bug, every issue that's
brought up very seriously. It gets straight to our engineering team and we try
to work as quickly as possible to resolve those issues so that people have a
really delightful experience.
And I think our early customer base has just done a wonderful job of being vocal
and helping us. They know we're a small startup and they know that there's a lot
of things we can do, a lot of things we have to say no to. I think it's just
great to have a community of folks that are using it at such a large scale and
then providing feedback to help continue to improve that.
Roland Siebelink: Absolutely. Very good. We are almost at the end of this
recording. What is something that you need from this audience? Are you primarily
looking for new employees? Are you looking for people to sign up and try it out
for themselves? What is the biggest need and where would you guide people that
aren't listening to this podcast and have become excited about Fireflies?
Krish Ramineni: Yeah. I think that we are definitely looking for folks to
join our team, especially on the go-to-market function. We're hoping to hire a
first growth PM and a more traditional product manager. So those are two areas
we're thinking about. We actually really like people from a B2C arena because
our product, the way we build and track and measure our product, is very much
like a B2C product, even though we're serving a B2B use case.
There are folks out there that have worked at fast-growing B2C startups or even
B2B SaaS companies that are bottoms-up focused, have a deep passion for growth,
metrics, and all of the things related to that, right? The full stack; you
should be curious about how growth works on a product level, how growth works
from marketing ads, PR, SEO.
The full growth stack, that's someone we're looking to hire for. And that is
probably something we'd love from the audience. If they're interested, they can
reach out to me directly or find me on LinkedIn, Krish Romineni.
The best feeling is when I join a meeting with someone and they tell me, "Hey, I
use Fireflies or I was in a meeting with someone else who used Fireflies." I've
had opportunities where I've interacted with colleagues of mine or friends of
mine, and they said I was in a meeting with someone they brought in Fireflies,
and this is a total stranger, you don't know them. We just spent 10 minutes
talking about how cool this is. For me, that makes all of the hard work - it
makes up for everything.
Those are the things, the moments I look for. If people are out there, they're
building teams, they're getting tired of having too many remote meetings and
need to better organize their meetings, we'd love for them to start using
Fireflies as well.
Roland Siebelink: Okay. URL is?
Krish Ramineni: You just can go to Fireflies.ai, just the name of our
company. It's two clicks to sign up and start using. And you can check it out
and test it out for yourself.
Roland Siebelink: That's awesome. And, of course, we will put all those
links in the show notes for this podcast as well. And if you know me personally,
and you'd like to be introduced to Krish for one of those go-to-market positions
that you mentioned, I'm very happy to introduce you.
So, with that, Krish, is there anything else we haven't covered that we should
absolutely still put in this podcast?
Krish Ramineni: I think we've gotten through everything. It was exciting and
I'm excited to see what people think about it.
Roland Siebelink: Awesome, okay. With that, thank you everyone for
listening. And if there's other founders of startups with traction that you see
as a good candidate to be interviewed on this podcast, drop me a line as well.
Thank you. And thank you, Krish.
Krish Ramineni: Great. Thank you so much, Roland.
Roland Siebelink talks all things tech startup and bring you interviews with tech cofounders
across the world.