Founders Interviews: JD Costa and Mário Tarouca of Framedrop

For this month’s Founders Interview we chatted to JD Costa and Mário Tarouca, the co-founders of Framedrop, one of the Pre-seed companies that was selected to join the 2023 batch of the Indico Founders Program.

Framedrop has developed a clip creation tool and platform that offers live streamers an easy way to monetize their content off-stream. Through their AI, they can identify a stream’s highlights and automatically clip them in a matter of minutes. For creators, this means no more long nights going through VODs or timestamping moments during the stream.

*This interview has been slightly edited for reading ease.

First, can you tell us about your professional background and the path that led you to Framedrop?

JD: I actually wanted to be something completely different when I was a young man. I wanted to be a fighter pilot for the Portuguese Air Force, something that I kind of achieved. I applied to the Portuguese Air Force, and I was selected, but during my time there I realized that there were some problems with the air force at the time. I noticed that they had some financial issues, and that the recently graduated pilots were not flying, or at least most of them were not flying, because the air force didn’t have the money to have that many planes in the air. So, for that reason and a couple of others, I decided to cancel that adventure, and what was left for me was university. I’d always liked sciences and ended up joining the biomedical engineering degree at the University of Coimbra. It’s a really cool degree, by the way, for anyone who likes sciences but doesn’t know the specialty that they want to follow. Biomedical engineering, just like mechanical engineering, is a degree that lets you study pretty much all areas of engineering, from math’s, to physics, to chemistry, to software engineering and so on. It was there that I learned basically everything that I know today, for instance, about artificial intelligence, which is something that we work with a lot at Framedrop. During my time in college, I heard about a thing called a Junior Enterprise, which I know Mário was also a part of. This was basically a group of students who essentially ran a company, inside the university, and each participating student had a role to fulfill. I found this very interesting because they did a few cool things related to engineering and at the time Coimbra was very prolific, in terms of entrepreneurship. This was around 2010, entrepreneurship was coming to Portugal as a whole, so a lot of initiatives were flourishing in the country, and Coimbra was the place to be. There was an institution called Instituto Pedro Nunes, which is an institution that supports startups and projects, they call it “technology transfers between universities and companies”. They helped our junior enterprise at the time organize a few events and it was then that I realized that there was this whole world of startups and a thriving tech scene. I learned about all that because of the Junior Enterprise. This one in particular was called Gknowledge, and it was absolutely instrumental for the path I followed after college because it got me started in this world. When I left college, I knew that I wanted to do two things. I wanted to do software engineering, which is one of the paths that you can follow if you study biomedical engineering, and I wanted to eventually start a company. I spent maybe eight or nine years working my way up as a software engineer, learning the ins and outs of tech startups, how a product team works, etc. Fast forward nine years, I felt that it was the time to start my own thing.

At that time, I had just started gaming again. I had stopped playing video games when I was in college because I didn’t have time and I had other interests, but at 28 years old I got back into gaming and bought a PlayStation. I had to learn how to use a controller for first person shooters, since I’d always used a keyboard to play FPS games before, and I started to watch YouTube videos to learn about how would be the best way to learn. It was at that time that I finally understood the content creation industry around gaming, which I previously only knew existed because my little cousins would watch people play games on YouTube. I’d ask them “Why are you watching others play instead of playing yourself?” My personal experience made me understand the power of this industry, and it was at that time as well that I started to explore opportunities in the industry. I understood that the live streaming industry was still very young — this was in 2019 — and the technology around live streaming was also still in its infancy. I knew that if you found good opportunities in a young industry, you should build a business. So, I first started playing around with a multicast app that would let you stream simultaneously to Twitch, Facebook and YouTube, but I realized very quickly that I had competitors far ahead who were doing much better things compared to what I had. Then, I saw this opportunity related to curating content from long form streams that can be five hours long, ten hours long, or sometimes even a day long. I thought in my head, “I might be able to build something around this, like an AI that is capable of processing the whole video from start to finish and select just the best moments”. I started building and all of a sudden, I had a proof of concept. I found a friend who was willing to invest in something like this, he said yes, and I quit my job to start working on what became Framedrop, at which point Mário joined me.

M: Since I was a young kid, I knew I wanted to create my own company. It was clear to me that when I arrived at university, I would be studying business. So, that’s what I did in Coimbra. During that time, I participated in several startup weekends, such as 3 Day Startup and those alike. This is where I met JD, 10 years ago more or less. I came across this awesome concept called Junior Enterprise, already mentioned by JD, and I was president of my junior enterprise, which was more so business focused, while JD’s was more engineering focused. When I finished my bachelor’s, I decided to go and try my luck in Brussels, where I applied, together with a team, to be part of the Executive Board of Junior Enterprise Europe. This is the European umbrella organization of all the junior enterprises across the continent. I was voted in and ended up spending one year in Brussels where 80% of my time was invested in talks and meetings with the European Parliament and Commission. So that you have an idea of how much was involved, I took more than 100 flights during that year across Europe. After that, I came back to Portugal, went to Lisbon, and I worked in several startups, like Landing.jobs, Unono and EatTasty. Actually, before working at EatTasty I set up my own startup called Breakfast, where we were delivering breakfast to Airbnbs. I spent two years there, having created a tech related startup without any developers. This is where I discovered business automation and no-code. Nowadays, I have my own course where I teach 12 no-code tools. So, to sum up, my background involves corporations, growth, and business automation. So, while I was still at EatTasty, which actually is also a part of Indico’s portfolio, JD sent me a message saying, “Hey, you want to exchange some ideas about the cool project I’m working on, Framedrop?” We spoke about different aspects of this potential business, and I started to become more and more interested, and five months later I was working full-time on Framedrop.

How is your product different from others in the market? What makes it unique?

M: So, there are a few differences. The first one is, if we look into our competitors being tools that finds highlights within streams, one first thing that comes up immediately is that we don’t require a sign up, or the fact that you need to download an app. With our competitors, you might need to sign up, create an account, download a desktop app, etc. In our case, anyone, at any time, any device can use our technology. The second thing to point out is that they focus solely on games. Whereas in our case, while we do build specifically for individual games, and we are adding more to the platform, we want to be agnostic. This means we allow anyone to process a stream with our algorithms, even if that stream isn’t related to a specific game. We are doing this already, we can already find some, what we call “funny moments”, or “sad moments” or “surprise moments” through different methods. Lastly, I would add to this list our business model. So, targeting streamers, we know that their purchasing power is not so large. They are not used to paying for tools or software. While our competitors mainly have a premium model, in our case, we believe that instead of charging creators — although we might introduce some sort of paid plan — we believe that we can actually make them earn more money. So, we have some ideas on how to offer these streamers a new revenue stream.

JD: We built our AI to be an external component that we can reuse and build projects around. So, we consider our AI our core technology. Since we started the company, we have been experimenting with products around this AI. We recently came to the realization that other people, third parties, might want to have access to this AI as well, even if they might have completely different use cases. Our use case is highlight detection and consequently reducing the time creators, and people in the industry like video editors, work on their videos. We want to streamline and facilitate their work, but there might be other companies that want to use our AI for other things. even our competitors might want to use our AI to do things similar to what we are doing, or to build new features. It might be less expensive for them to license an AI through an API, than developing it internally. So, this is something that we are starting to explore. Even though we don’t have this working yet, it might differentiate us from the others even further in the future.

You are still a relatively young company, but do you have any tips or advice for someone who is embarking on their entrepreneurial journey?

JD: If you’re starting this journey, even if you have some cash from family and friends, or even a VC, be careful and try to keep your team structure very lean until you find product market fit. At early-stage companies like Framedrop, if you feel like you don’t have product market fit yet, you will want to experiment a lot with very diverse things. So, you might want to build a product today and build another one in one month from today that is completely different. The team that is required to build product A might be very different from the team that is required to build product B. I’ll give you an example. You might want to build a product that is very demanding on engineering, and you will try to hire a team around engineering. You’ll need an AI engineer, a back-end engineer, a front-end engineer, a designer, and so on. Then next month, you realize that you have to pivot to something very different and that the new demands are around sales and marketing. All of a sudden, the team that you just hired is not the team that you need. That’s a very bad place to be because you’ll have to make some hard decisions. I’m not saying that we had this problem. Fortunately, we feel that we have a very strong core team that is very flexible in terms of structure and in terms of mindset. This allows us to change things very quickly. A good example of this is that we pivoted recently, like one month ago. We realized by the end of the quarter that we were able to build a completely different product in just one month. That’s very powerful, because the worst thing that you can do before you have product market fit is to be hard-headed and keep pursuing something that you’ve already realized will not go anywhere. This is my one piece of advice: keep your structure lean. If you feel that you need specialized work, look into hiring freelancers instead of hiring someone full-time, because you might not have work for that person in the next quarter.

M: I agree with everything that JD mentioned, I would just add one thing. At an early stage, what you want to try to do is to create a minimum viable product to showcase to your potential customers. You might not even know who they are yet, or even in which industry they are, but my main piece of advice would be to always think of a way to try to create this MVP without code. Again, I’m a no-code lover, but there’s always a way to create this MVP. For instance, at EatTasty, and I wasn’t there at that time, so I’m not biased, what they did at the beginning instead of creating an app or a website for people to order their meals, was a WhatsApp group. So, they onboarded some people from Lisbon as initial clients, and every day they would send a message announcing the plates they had, and people would order if they wanted. This allowed them to test their product market fit with significantly less associated costs.

If you could go back in time, is there anything you’d do differently?

M: As you mentioned, we are still young, but we already have some learnings. I’m going to mention one related to recruiting. If you as a founder find some red flags about someone, even during the recruitment process, that might not be the person you want to bring to your team. At least, if it’s possible, you should do an experimental trial of a few months. This is because when you have a team of a few people, one person will have a lot of weight on the processes and culture of the company. So, I agree with what JD said, which is keep your team lean and small. Adding onto that, pay even more attention to the ones that you bring into your core team, because if it goes wrong, you will pay really hard later on. By the way, if you find a person and it ends up not going so well, don’t try to postpone asking that person to leave. It’s tough but in the long run you will thank yourself.

JD: I have another one: keep working on product during fundraising. For context, we built an MVP and it worked, although it wasn’t final. We were getting a lot of feedback from creators and we were approaching a time where we were running out of money from our first investor and we realized that we had essentially two options. One of them was to keep working on Framedrop the side — so Mário would get a job, I would get a job, and we would work on Framedrop during nights and weekends — and we agreed together that that wouldn’t be the best option, because both of us are very dedicated to what we do and it would be very hard to do both things at the same time. The second option was fundraising enough money to put together a team and keep working on the product. So, the idea of raising the money was to build product and validate the product. At least that was our pitch for our first funding round. One mistake that we made is that the fundraising process was very, very time-consuming, both for me and for Mário. I have the perspective of the product person, or the engineering person here, Mário from the operations and marketing side of things, but we were both co-founders. We were both very involved and we wanted to be very involved in the fundraising process. So, we essentially did this work together. Every decision we would make together, every call with potential investors we would have together, every bullet point in the term sheet was discussed together, and the consequence of that was product development just stopped. We stopped product development entirely for I would say seven or eight months, which is the time we took to close the funding round. Also, while we’re here, another piece of advice, don’t fundraise during the summer, because people won’t answer, they won’t be available, and any documents you need signed you might not get until September. So don’t do your fundraising during the summer. That’s a little detour, but it’s important. So, if you have a small team, product development just stops, and eight months later we found ourselves looking at our product and realizing that our competitors had advanced with features that we did not have yet. Some of our assumptions were already wrong because the industry evolved very quickly, and it wasn’t a very good place to be during those months because we had to reconfigure some of the ideas that we had in mind in terms of product. So, to summarize: while you’re doing fundraising organize between founders to split the work and at least one day a week, two days a week, keep working on product, keep talking with your potential customers, keep getting feedback. Don’t stop doing it because you will be late to the party when you’re back to product development.

What has been the role of Indico on Framedrop’s journey?

JD: So first of all, Indico was a very pleasant surprise, not because we thought that it was going to be bad, but because we didn’t think that it was going to be this good. Indico has been very close to us and very active. Indico has been proactive in talking with us, recommending things, posing doubts, which is really good, and at the same time Indico has not been too intrusive, which was one of the anxieties that we had going into fundraising. This is the first time I’ve dealt with investors directly. I’ve been in companies that were VC funded, but I had never dealt with investors, I had never been in a board meeting. I know that Mário had, but we had this anxiety about what would be expected from us as soon as we signed the deal. The experience with Indico has not been intrusive at all, we don’t feel negative pressure to do things and to advance, they are giving us space. Also, something that I feel, I’ve actually never discussed this specifically with Mário, but I feel like Indico knows exactly the early stage that we are at, and they are giving us space while trying to help us to get to the next stage. They are not expecting us to advance before we are prepared to do so. They don’t expect us to grow in terms of metrics because they know that we are still finding our product market fit for instance, and they try to help us find our product market fit. I believe that that’s the best thing to have from an investor, almost as if Indico is kind of like a wingman for us. We’ve had bi-monthly calls lately with Cristina, and probably because Cristina was a founder herself, she already knows all the problems that we bring up in those meetings and she’s been very helpful in telling us what we need to hear, and we’ll usually use her feedback and apply her feedback. We feel like she has really helped things keep rolling smoothly. That is very cool, I’m very happy with Indico and Cristina’s specifically, the interaction has been awesome.

M: Just to add one or two things here. I completely agree with what JD said, specifically regarding Cristina. I knew her from before, but having her in our meetings, knowing that she already went through what we are going through now, gives us not only strength and belief, but extremely valuable knowledge. Another point I’d add is, although we talk a lot about operational things, we also believe that you guys put a slight but very positive pressure. When for instance, we talk about where we want to be in six months or one year from now, you also sometimes take us from the operational day to day do this sort of helicopter view of “where do we want to get to?”

Lastly, what has been the most challenging part of growing a startup? And the most fun part?

JD: For me, the most challenging part so far has been finding product market fit. The theory is out there, you can read many books about it, you can read many online publications, you can listen to talks, experiences and testimonials from other founders. You will want to try many things, most of them recommended by what you read and what you hear, and still not get there. This is something that for me, I see more like an art, because you can take all the right steps and still not get there. It’s very difficult because of that. There are so many variables that go into defining product market fit, so, I think that you only have product market fit when people step up and ask to pay for what you are offering. This is very, very challenging. You might be lucky and hit the target with your first shot, but if that doesn’t happen, it can be a very painful process for the founders.

In contrast to that, what’s been the most fun part?

JD: Fundraising was really fun for me, but a bit painful because it was a new thing for me. I’m an engineer. I’m used to solving problems by being very practical and pragmatic, but fundraising is a completely different beast. It’s not pragmatic at all. If you’ve been there, you know. It has to do with selling a dream, telling a story, trying to convince people that you are to trust you. Moreover, it’s not about convincing people separately, you have to convince person A, B and C, and then convince them to work together to help you set up your round. That’s a complex puzzle. Also, will your investors fit together? It’s very much a people’s game and you have to become a people’s person in order to manage all of it. So, this was very fun, completely different compared to the problems I am used to solving.

M: Its somewhat similar to what JD said, but from my side, is the challenge that there is no right answer. You can talk with experts, talk with founders, get feedback from your users, analyze all the data that you have, but in the end, its on the founders to decide where to invest their resources. Its only after a few months, and sometimes even years, that you find out whether the decision you made was right or wrong. Even then, you won’t know if that was the most optimal decision. So, the challenge is that there is no right answer, but on the other hand, that’s the cool part of it, it’s a challenge. If you are open to it, if you want to be challenged every single day, then that’s the most fun part. We started with nothing, you know, and now with the great team that we have, we are able to believe that we will be able to create a product that will bring value to its users and solve real problems.

We hope you enjoyed this interview! JD and Mário also suggested some valuable books that they say first time founders should definitely read:

• Traction by Gabriel Weinberg and Justin Mares
• The Great CEO Within by Alex MacCaw, Matt Mochary, and Misha Talavera
• The Mom Test by Rob Fitzpatrick

*Interview by Álvaro Furtado