Sara Lindquist:

I am Sara Lindquist from FUSE. We're an early stage venture firm based right here in the Pacific Northwest. And just like the founders in our portfolio, we are just getting started. We believe that founders deserve more, more urgency, more community, more expertise, more reliability, more of everything, and we aim to deliver. Join me as I introduce each of our portfolio companies in the FUSE family to date.

Today you'll hear from Benji Brash, Co-Founder and CEO of Roboto. Join us to learn about Benji's experience helping run the drone program at Amazon, and how it paved the way for starting Roboto. This business is transforming the way engineers can access and analyze multimodal hardware and sensor data. Let's get started.

Benji, thank you so much for being here today and sitting down with me.

----------------

Benji:

Yeah, real pleasure. Thanks for having me on. This is awesome.

Sara:

Of course. Yes, virtually sitting down, of course, but here we are. We're both sitting and looking at each other through the video camera, basically the same thing. But anyway, it's so great to have you here and this is going to be a super fun conversation. So to kick things off, I would love to just dive in and have you share a quick bit of background on what Roboto is and how it's serving customers.

Benji:

Roboto makes it easy for anyone really that's working in robotics to search and analyze all of the sensor and logs that robots produce. And robots, when they're operating, produce huge amounts of data, could be gigabytes, terabytes of data in just a few minutes of operation. And all of that data is multimodal, we say. So it could be images, it could be point clouds from a LIDAR, could be lots of time series data, and robotics really designed to work with that data from the ground up and leverage the latest in AI to do it efficiently.

Sara:

Yeah, love it. That's awesome. And I would love to dive into the space more. And to start, I'll have you walk us back in time a little bit. I certainly know about your background, but I would love for you to share about your past experiences, the Amazon Journey, AI two experience, and then what was really the light bulb moment for you where you said, "Hey, time to jump ship and go build this thing"?

Benji:

Yeah, totally. So me and my co-founder, Eve, we both worked at Amazon for I think about seven years or so, and that was really in robotics. And the types of robotics projects we were working on were drone delivery. Amazon Prime App, one of the really interesting flagship robotics programs that Amazon's been working on for several years. And that was an amazing time to be getting into robotics just after I'd graduated from school. It was kind of an amazing job to start with where you're really building innovative technology that's sort of not really been done before. It was a baptism by fire, I think, into that robotics space.

Amazon has really been a leader in robotics for a long time, but drone delivery was certainly to the next phase of that investment for them as a company. And really over time, I worked in building sensor related products, so optimizing kind of all the data that was coming off of the drones. And also autonomy algorithms, how we were kind of able to fly safely and make sure that everything was robust and reliable. And later in my Amazon career, I worked on simulation too, and that was a really interesting place to work. Because that was sort of the way that we built confidence in the system before we went out and tried to fly drones in real life.

Sara:

That's so cool. So cool.

Benji:

That was really almost sci-fi, building a large scale simulation environment and saying, "Hey, if this crazy thing happened, what would the drone do in this situation? And I think that gave me a really interesting appreciation for what it takes to build robots that need to be safe and reliable and work in the real world. And similarly, my co-founder Eve came from a computer vision and machine learning background. So he was really deeply involved in making those vehicles safe and reliable and autonomous ultimately.

And so I think at some point, we both were exchanging notes and we've been friends for a long time. And we both noticed that we were getting very slowed down working in robotics because of the amount of data that these systems were generating. And we kind of looked around the organization and we found that really everyone was being slowed down by the amount of data that got generated. And just to put it in perspective, we would do a drone flight of some kind. As I'd said before, it might generate gigabytes, potentially terabytes of data. And if something went wrong, we'd need to investigate. And you can imagine what it's like to pour through gigabytes and terabytes of data.

It doesn't sound fun and it isn't fun.

Sara:

Yes.

Benji:

Exactly.

Sara:

Woof is my response there.

Benji:

So you need to be really smart about how you use that time, and sometimes you're just sitting there waiting for all the data to download as well, which sucks. And so we were both frustrated by that, and Eve was trying to train new computer vision algorithms, and he would always need to get just a subset of the data that we were collecting to use that to train new types of algorithms. And that was a real pain for him too. So we looked at each other and we were like, we think there's something we can do about this. We started doing some validation outside of Amazon and making sure that it wasn't just a problem that we had uniquely faced.

And so we talked to friends and that were working in other robotics companies at that point, and discovered that it was even worse for them, because they didn't have the resources of an Amazon behind them. And we've really found that in discovery now that larger companies have actually a lot of resources can afford to hire lots of backend engineers, infrastructure engineers to process this kind of data. They still maybe aren't doing such a good job of it, but at least they can hire people to try and solve it. Whereas small and medium sized startups and robotics companies really, really struggle with it. It's a huge pain point because they just can't necessarily afford the headcount to bring in a big infrastructure team.

Sara:

Well, and I was going to ask you that. You spoke to it at Amazon, but how are people even attempting to do this? It's headcount or bust is kind of what it sounds like.

Benji:

Yeah, pretty much. And that's really sad right now that there just really wasn't and still isn't a good off-the-shelf offering that companies can just buy. There's some amazing data platforms out there that work well with what we would call traditional data, business data, kind of tabular time series data, but as soon as it's multimodal and you've got images, LIDAR point clouds, radar data, just these systems were never designed to work with that kind of data. So yeah, bust.

Sara:

Okay. And so you said, "Why not us to go do this?"

Benji:

Yeah. I think we know what we would've needed. We felt the pain, and we really kind of confirmed that when we went out and talked to other robotics companies and other robotics engineers. So we felt like there was a big opportunity there, especially with some of the recent advances in AI to accelerate that too. And longer term, what really excited us was that we see robotics as really the first industry that's kind of living in the future in some way. Robots are already generating a lot of this multimodal sensor data, but lots of other industries are going to be generating this data too. Even industries like aerospace, manufacturing, they're starting to generate a lot of multimodal data because they can integrate cameras and other types of sensors more easily than ever before. So really, we see kind of the long term, it's just all of the sensor data needs to be wrangled somehow and wrangled very efficiently.

Sara:

Absolutely. No, it's a really exciting opportunity and it feels endless in so many ways. So just kind of circling on the team, you and Eve. And is the team bigger? Have you added some more people? Yeah. Okay. Tell us about that.

Benji:

As of today, we're a team of five. One hour ago we signed a really amazing backend engineer who's joining us, so we're stoked about that.

Sara:

Oh, fantastic. Okay. Five of you. So now that you're growing the crew on the ship, I would love to get your perspective as a leader. What's your perspective on the culture trying to build at Roboto, and why do you think that your team is the place where people should come?

Benji:

Yeah. Yeah, I think it's a really good question. It's a hard one, I guess. Ultimately for us, we're all really passionate about robotics. I think that's kind of a really key part of our culture. We all think that robots are really going to be crucial in society in the next decade, with labor shortages, aging population, even food shortages. I think really robots are the answer to a lot of these problems. And a lot of companies building robots are still held back and finding it hard to get into production, because of issues like not having scalable data infrastructure and AI tools that make it possible to get through all that data. So I think we're really aligned on that vision of how important robots are going to be in the next decade. And then we're also just generally very passionate technologists, and we really love getting our hands dirty, using the latest to create large language models to figure things out. So I think that's the culture we're building right now.

Sara:

Well, it is a cool blend, because you have people, I mean quite literally, getting your hands dirty, going and actually having the experience in robotics, but then just being able to just alleviate the bespoke nature of what you actually do with that data. So it's cool, well-rounded team you're building. So my next question for you then is just about the startup journey as a whole. You've experienced this pain point a lot in your past roles, and have a ton of experience and have a clear mission ahead of you. But I'm curious, what continues to keep you inspired? Startups are never linear, there's lots of ups and downs, so what is kind of the key thing that keeps you moving forward and keeps you inspired?

Benji:

For me, it's really the opportunity at a startup to wear a lot of hats. I think that is so stimulating. And I'd started my career at Amazon as a software engineer, and I really love being a software engineer. It's very satisfying, really solving puzzles all the time, especially working in robotics. It's so physical and practical and tangible. You really see what you're building is a thing. You can interact with it. And maybe you don't want to get too close to it because it's a robot sometimes.

Sara:

This things going rogue.

Benji:

Yeah. Ultimately, I feel now at a startup, we're just able to really control our own destiny and focus on things that we're really passionate about and wear a lot of hats in the process. And I feel in the last year, I've just gained so many new skills that I'm really thankful for. And just the people that I've met in the startup community, in the startup industry as well are so generous with their time and advice too, that I feel like you just suddenly tap into this network and community really that all are kind of cheerleading for each other. It's really cool.

Sara:

Oh, I love that. Good answer. Okay. One totally off the cuff, what's actually the craziest thing that you've seen a robot do that you've worked on? I'm sure you have some really good stories here.

Benji:

Yeah, I need to be careful probably with what I share.

Sara:

Okay. If you have any shareable responses, I'll take it.

Benji:

Yeah. I think the amazing thing when you're working in robotics in unstructured environments, is that the world is just really a random number generator. Anything can go wrong. And certainly when we were testing robots in outdoor environments, you never expect bird poop on a camera to cause you problems, right?

Sara:

For sure. I'm sure that causes complete havoc in some cases.

Benji:

And also just the weather. I think that was something I've always taken for granted, especially as a software engineer. My background isn't certainly hardware engineering, but just building systems that are really kind of robust and weatherized and can work in an environment that's always changing. Might be hailing one minute and sunny the next minute, is really complicated.

Sara:

100%. Oh, that's fun. Well, maybe we can get some more stories over a beer one day or something. Not in a public forum.

Benji:

Absolutely, yeah.

Sara:

Yeah. Anyway. Well, that's awesome. Okay. So in closing here, I have one final question for you. This is kind of the call to action to those listening in. So what right now, do you need more of? How can anyone get involved or help?

Benji:

Well, like I said, we're really actively hiring. If people are passionate about the robotics space and interested in AI and using AI to get through massive amounts of data efficiently, then please hit me up. We're hiring folks on the front end side, backend side, and also AI machine learning. We've also just released our sandbox environment. So the sandbox is really a kind of first version of the platform that we're building, and it's loaded up with an example dataset. So I'd love to encourage anyone to go to roboto.ai, check out our sandbox, play with it, and experience the magic of really using natural language to search these giant automotive datasets in the sandbox's case. And that obviously, finally, if anyone's working at a robotics company and feeling some of the pain that me and Eve had felt and spending more of your day wrangling data instead of doing algorithmic work or real engineering work, get in touch. We'd love to talk to you and show you what we're doing.

Sara:

Yes. That's fantastic. Well, Benji, thank you so much. This was a really fun conversation and we're just all really excited about what you're building and all hands on deck over here to help make this successful. So thanks for taking the time.

Benji:

Thank you so much and really appreciate it. The whole FUSE team's been awesome, and we're really excited to be on this journey together. So thank you.

Sara:

Well, it's going to be a lot of fun.

Benji:

Oh, yeah.

-------------------

Sara:

If you want to join a stellar team that's working to completely alleviate the bespoke nature of complex data analysis, be sure to reach out to Benji directly. And as he said, try out the product. All feedback is welcome. Thanks for joining us, and we'll see you on the next one!