Can AI and BigData change the way we feed the world? XpertSea believes they can.
An interview with XpertSea CEO, Valérie Robitaille, on the fish-counting device that could help aquaculture become the most efficient protein source in the world.
The beauty of technological innovation is that it can transform industries that have been working a certain way for decades or even hundreds of years. While shrimp and fish farming can be traced back as far as the 15th century in Indonesia — where rice paddies were converted to brackish ponds during the dry season — industrialization only began in the 1930s and the farming methods haven’t changed much in the past hundred years.
Among the greatest problems for hatcheries has been the necessity to hand count eggs and larvae. These rough estimates make it difficult to manage production — a problem that is now being addressed by Québec City-based XpertSea. Using the company’s XperCount device, fish farmers can accurately quantify the number of animals, measure their size, assess growth and mortality rates, and provide forecasts after taking multiple samples. Having announced their $10 million Series A this April, we spoke with co-founder and CEO, Valérie Robitaille about XpertSea’s product, entrepreneurial journey, and how the company discovered their mission to help feed the world in a more sustainable way.
Can you please start by explaining what XpertSea does?
Aquaculture is the fastest growing food producing industry in the world. It also has the potential to be the most efficient source of protein production. While it has a lot of potential, in practice there are still big technology gaps. There’s not a lot of technology available to characterize the quality of all the little organisms, so there are a lot of problems in managing the production. There are also a lot of losses and disease. XpertSea technology allows producers to count, size and image aquatic production. All of that information goes into our cloud-based platform and we add additional intelligence to help producers to optimize their feed and optimize the growth of their organisms to make sure that they get the best results.
How did you get into this industry? It doesn’t sound like something that someone would just stumble into.
My background is in marine biology — marine science. That’s what I did for my undergrad in the US in Maine. I came back to Canada for grad school and I worked with optics and photonics in the marine environment and I started to have lots of ideas for commercial applications. I had met a mechanical engineer in Maine and together we started building little prototypes for characterizing the ocean with optics. We wrote articles that were published and a company in aquaculture contacted us to see if we could use [our device] to count shrimp larvae. We didn’t think too much of it at the time, but then we discovered it was a huge market with big pain points.
You say that you saw a lot of opportunity in the market. Are there any other companies doing similar things?
We’re really specialized at the early stage — the beginning of the value chain — at the hatcheries. That’s where the eggs hatch, and where they grow the larvae and organisms to a particular size before shipping them to farms. For hatcheries, there weren’t and still aren’t really any commercial solutions for the quality control of production. We’re the only one — competing with the traditional hand-counting methods they’ve been using for decades.
How does hand counting work? Would they simply take a bucket for a sample from the tank, count what’s in it and then estimate what would be in the rest?
Something like that. One way is they would count a thousand, put it in a little bowl and then just visually try to pour a thousand in other bowls based on the one that’s counted. There are a lot of techniques but all of them are pretty inaccurate.
Can you give me a little more detail about how your system works? Are you also looking at small samples or are you getting more visibility into the whole tank?
Our current product is a 5-gallon container. If you’re doing shrimp eggs, you can easily count a million and then stock a million in the tank. You can also track the growth of the organisms by taking samples over time intervals. Based on the number of samples you take and the variation between the samples, a statistical tool in the app tells you the margin of error so you can keep sampling until it’s at no more than five percent. Our goal isn’t to just sell the device but to really provide a solution that helps farmers improve their process overall.
Are there other things that are helpful to understand your product?
We use camera optics. We’ve integrated a lot of machine learning, training frameworks. We have a backlog of about 80 different species that people want us to create applications for. What we do is use the same hardware and develop different apps based on the species.
So you’d be able to branch out into a lot of different species then, based on what people are looking for?
Exactly. We even just did an application for insects. Anything that’s small and hard to count or manipulate, we can help with!
Was entrepreneurship always something you were interested in or did you consider research before you thought about creating products?
Since I was little, I’ve always had an entrepreneurial side. My grandad started a big company in Québec and when I was a little kid I would draw magazines about dolphins and send my little brother door to door to try to sell them in the neighbourhood.
When I was in grad school I started giving consultation services to engineering firms. I would rent out the university lab and take contracts to do environmental surveys. Then I started to have the idea for a product… and when we felt this idea had a lot of potential and we needed to get it to the next stage, we raised funding and moved toward XpertSea as we know it today. I liked research but I always felt that it took too long to have a concrete impact on the world. That’s why I was more comfortable with business and having products that could be launched quicker.
Can you tell me more about your team?
The engineer that I started the product development with is actually my husband [Cody Andrews]— who I met in Maine. And quickly my mom [Sylvie Lavigne] joined the company because she was a VP at an engineering company and had the experience we needed to manage suppliers for building our product. My brother [François Robitaille] is a CPA, so he also joined when we decided to finance the company. It was kind of a family affair at the beginning — it just happened that I had people who had complementary skills who were much needed and ready to take a big pay cut to start the project. Since then we’ve had a lot of really great people join the team.
How big is your team at the moment?
We’re a little over 30 people now.
And how much do you plan to grow?
The goal is to double in size by the end of 2019.
What’s it like having your family members as your co-founders?
I think it’s a huge advantage. It’s really hard running a startup! You go through so many ups and downs and highs and lows and having people you can trust 100 percent — for us it worked out really well. What’s actually harder is the perception of others: thinking maybe we don’t all have the necessary skills. We know we make a good team and have complementary skills but we always have to prove that it’s the right match and that we’re willing to bring in the right people with more experience.
What have been the highlights of your entrepreneurial journey since you started XpertSea?
Right before we decided to look for funding, I went to India with my brother and we decided to organize a seminar. We walked in and there were a hundred Indian farmers in the room in a very small village. We were there presenting a pretty high-tech product and when we finished the presentation everybody started asking questions and they were super engaged. That’s when we realized there’s really interest for this — there’s really a market. At the time, all of the buckets were hand-made. It would take us days to make just one. So Frank, my brother, being from the financial side, said: “Let’s raise some money and do this thing right. Let’s go capture that opportunity!” It was the moment that we realized that we really had a venture in our hands.
How about a terrible experience that you’ve learned from?
We went to Vietnam to do a live demo for a pretty big Chinese group. There were about thirty people and they all had their cellphones out, super excited that they were going to put some shrimp into the bucket and it was going to count them. Everyone was recording and they put about 2000 in there. Our device has a countdown… and it was like the countdown of death. We were so stressed… at zero it has to give the right number. So it does the countdown… 5…4…3…2…1… and then it shows 7000. It was a horrible feeling.
We actually have a picture from that demo in our production office to always remind us to be prepared. The early days were very hard because when we developed the product we did not account for all of the different use-cases so things would go wrong. I’m glad those days are over now.
Do you have any advice you’d give other tech founders? Anything you wish someone had told you?
One important thing is company culture. For us, it’s super important and now we really see the payback of having our employees super engaged. I would say never underestimate [culture] and make it a priority in the company. It’s super hard to find talent and keep talent and this is something you have to fight for, especially as you hire more experienced people and people from different backgrounds. It pays off if you keep it a priority.
What are your plans now that you’ve raised your Series A?
Right now we have a very good product market fit in shrimp and we’ve proven that we can sell the product [in that industry]. So a big part of the funding will go to solidifying our go-to-market strategy in the shrimp industry — make sure we have our feet on the ground, and make sure that we can keep building traction. We will also use it to increase our technological capabilities because we want to expand horizontally into different species but also vertically in terms of more data products. We need to grow our data science, machine learning and software engineering team as well.
Could you speak a little about your fundraising process?
It was very good and very stressful. I’m glad that it’s done. We wanted to find investors that had industry knowledge plus tech contacts and knowledge. It took a while but we didn’t settle and we were able to find the right partners. We could have optimized the deal by 10–20 percent but having the best partner is definitely the way to go.
Are there things about the preparation process that really helped you?
I did a lot of the pitching and I think it really makes a difference if you really know everything inside out — especially for your market and your customers. It seems to make the investors more confident so I would say to founders that even if you have great, experienced people in your team, or other people who might seem better suited to pitch or talk with investors on paper, make sure that you go with either a founder or someone who really knows the business well. In the end, that’s what seems to make investors feel more secure — investing in someone that knows the opportunity more than anybody else.
Is there anything else you’d want to add that we haven’t covered?
We were very lucky to have very good, supportive investors from the start. It’s not always easy — everybody has different goals — but with Real Ventures and some of our early investors we were so lucky. They were really supportive and helped us raise that second round. Sometimes we hear things [from other entrepreneurs] and it doesn’t always work out so well.
It’s also cool that people are talking a lot more about aquaculture now. We see a lot more in the news, so hopefully it will keep going that way… and that we’ll have something to do with its growth!
This interview was conducted by Lauren Jane Heller. It was edited for length and clarity.
Follow us for more in-depth interviews and insights from tech entrepreneurs.