Matthew Youngblut - founder at flockbio.com

Welcome to another Founder Wisdom Podcast. We have Matthew Youngblood with us today. He's the co-founder and CSO, Chief Scientific Officer at Flock Bio in Alameda, California. Interesting company, interesting guy. Really dug the concept. He's the next Ginkgo. He's the next Twist. So he's got the experience. And I believe your co-founder is pretty high pedigree as well. And he's also named Matt. So Matt, welcome to the pod. Tell us a bit more about yourself and what you've been up to.

Yeah, so I'm Matthew, CSO, co-founder here at Flock Bio. We started the company a little over a year ago because, as you mentioned recently, we were a previous acquisition of the biotech company Ginkgo Bioworks. They've since divested a lot of those previous acquisitions that they made over the pandemic, and so we basically just kind of picked up the lab, moved down the street, and opened up the company that we always really wanted to have because we really, we specialize in high-throughput molecular biology in order to be able to run experiments where you can test, you know, thousands, millions, or even billions of questions at the same time. But a lot of those ways to go about that are really bespoke and really custom, and there's not a lot of providers out there in the world that really, you know, offer these products. They offer them in more of like off-the-shelf or somewhat customized, but the highly customized work usually falls on the researcher themselves. And, you know, the thing these days, you know, with so much with AI and folks being able to empower themselves, you know, my big, you know, kind of saying these days is scientists should focus on what they do best and partner for the rest.

Tell me, let's start first with your background, right? Like, I knew about Ginkgo, I didn't know about Twist. Let's start with Twist. How was your time there? What did you learn as the manager of R&D?

Yeah, so Twist started about 12, 13 years ago in the Bay Area here as well, focusing on miniaturizing synthetic DNA production. And they've really become one of the industry leaders on this front. And I joined the company at the beginning of COVID at the height of the pandemic there in late 2020, when there was that huge push in biotech. A lot of that kind of went away because of, you know, the ebb and flow of the pandemic, but we're seeing a huge resurgence in just the last couple of months. I think that's really going to carry through in the next few years of this new wave of biotech, and Twist is a company that really enables that because of their scale. They're able to do a lot of things at a price point that wasn't really possible before.

Interesting. And what about Ginkgo? Ginkgo is like less on a scaling path, but they still IPO'd, right? What did you learn from your time there?

Yeah, so Ginkgo had one of the largest biotech IPOs ever in 2021 with $1.6 billion raised there. The thing there is, you know, the real ethos is, you know, making biology easier to engineer. Well, biology itself, you know, it's evolved for billions of years. It wants to do its own thing. It doesn't really want to be engineered. And so the amount of things that they took on all at once were just kind of beyond, you know, kind of like scientific feasibility. And so the company has since kind of refocused in some key areas like high-throughput automation, things like that. But Yeah, that was the real thing there is like the company, you know, took all that money and really tried to grow as much as like beyond what was physically possible. And that was just not sustainable.

Interesting. And what led you to actually launch your own company? It's quite uncommon as someone in R&D, as a scientist, right? Like what was the spark, the ignition?

Yeah, so I've always had this kind of entrepreneur kick. I've had, you know, various businesses here and there kind of on the side along ways, you know, in IT or things like that, helping out other companies. And so it's always been on my roadmap to start a company, and I've been wanting to do it for years. Previously, before Twist, I was CTO of a small startup called NextBiotics, working on phage engineering for animal health and agriculture. And I really loved, like, the startup, you know, realm and everything like that. But at the beginning of the pandemic, I left startup land because my wife and I had our first child and things like that. And it's not super compatible to have, you know, young kids and things like that in startup land. I'm really impressed by folks that can make that work. And so I went back into more corporate science. I was at Roche previously in large corporate science. So I was at Twist, I was at Ginkgo, but when Ginkgo, you know, closed the lab and everything like that, had the opportunity to find the next thing and— Because we had severance and everything, it gave me that opportunity to explore some options. Somehow I convinced my wife that this was an okay thing to do with a 3 and 5-year-old at home. But, you know, with that support and everything and the network that I've been able to establish with Ginkgo and my co-founder that I've worked with for years has really allowed us to kind of hit the ground running. If we were starting from scratch, this would be This would've already failed.

Interesting. And explain the concept again for non-technical folks. Like, what do you do exactly and why this is valuable? Seems it's like a hybrid of consulting slash agency and software, right? It's like bridging the gap between these scientific companies and with the data, right? Like, it would be complex for them to get all that data. Tell us a bit more again what you do.

Absolutely. So yeah, we really do bridge that gap between DNA synthesis and the experiments that folks are trying to run at scale. You know, there are a lot of ways to go about building these types of experiments and custom DNA. What we do is we work with a wide range of DNA synthesis partners throughout the country that produce a huge range of different products. They have different advantages and disadvantages, and we're really power users of these types of products. So we know how to get them to all fit together and build the experiment or the product that the customer really needs without making compromises that would normally be necessary just to get the parts to fit together.

Interesting. How do you get your clients? Is it like your personal network, your co-founders? Like, what's the GTM play here?

Yeah, so the number one go-to-market is social media. LinkedIn has been our number one, you know, outreach and we've been most successful with. I, you know, the email cold outreach that we've tried, you know, so far with like Instantly and things like that haven't been super successful. But I think you've got some areas on that front that could perhaps help us out. Other than that, yeah, the network that we built up between Ginkgo and all these clients and everything over the years, And who's the ICP?

Is it like CSO? Can you define like the ICP precisely?

Yeah, that can really depend on the company size that we're working with. So typically if it's less than like 30 to 50, then it's pretty much everyone's at play, whether that's, you know, top leadership or, you know, heads of, you know, heads of R&D, even like director or even like, you know, folks, scientists, RAs at the bench too can be really good initial points of contact at small companies. At large companies, especially large pharma, like you don't want to go to the CEO of Novartis. They're just gonna be like, well, that's several layers below me. And so there, that can actually be fairly tricky to pull out which groups within those giant umbrella corporations to actually target.

And Novartis, how would they treat these libraries of data? Would they already have a supplier or have someone in that role that fully manages it? Would they be a good client for you or too big to establish?

Yeah, yeah, they would be a great client. The thing with that is they most likely have, you know, established channels and things for the workflows that are in motion. But one of the things that we offer is speed. One of the main value props that I communicate to our customers is time, because instead of, you know, working through the same old issues that exist, you know, along the way, we help our customers leapfrog over that and just get to the data that they need. And so large pharma, they might already have like, you know, built-out workflows, but if those workflows are you know, take a good amount of time. They're always looking to speed things up.

Interesting. And you deliver these faster. Is there a cost saving on my end? Say I buy the data library from my end, like, would I pay less? What's the advantage of working with you versus getting these myself?

It, uh, it's, it's a little situational. It depends on scale. A lot of times we are— we come in at a lower price point than other established players. There are other times where, you know, we're a little bit higher in cost because of the extra steps that we go through to be able to help our clients. I would say we're not just a black hole that folks drop a spreadsheet into and hope they get back something that works. We work with our customers to be able to understand their upstream and downstream needs. So that way we, we don't just build a thing, we build them a thing that actually works for them.

Got it, got it. Is there some compliance stuff for your company that you need to comply with, with all that data? Tell me a bit more.

So one of the main concerns there is biosecurity and, you know, that folks, especially with AI these days, they could kind of go off the rails and order, you know, do things nefarious. And so we have incorporated the IBIS, I-B-B-I-S, biosecurity screening COMIC tool into our intake process, which actually goes through in this— IBIS is a nonprofit located in Europe that focuses on generating biosecurity tools for the industry in a free way. And so we've been able to implement that and incorporate that very easily so that way we can certify that every sequence that we build passes those checks.

And that data, what do they do specifically with it? They launch new experiments? Do I get something physical as well? Is it only online and a CSV sort of?

So we actually, we're not collecting like the final assay data here at Flock. We actually have a set of partnerships that we've built up to be able to do that more downstream work, whether that's in cells or mice or NHPs, things like that. And so what we offer is the customized DNA that contains, you know, the thousands or billions of sequences in that, along with the sequencing data and QC data to certify what's in the tube and that it is assembled correctly and that it's what the customer wants.

Got it. And what do they do with that data? Let's say I buy that data, right? Like, is it to run some of my experiments? Is it to push it against some of my results? Like, what do I do with this data?

Yeah, so most of the time customers are taking the DNA and then they're putting that into either cells or there's a lot of cell-free protein synthesis work that's done these days, which goes, you know, really quickly. And then people— we're building these libraries for ultimate next-generation sequencing readout. A lot of times we do barcoded libraries, which are short stretches of DNA that are unique identifiers that we'll associate with a variant with long-read sequencing. And then a customer, all they have to do is just sequence that short bit, which is relatively cheap and scalable. And then what they do with that is they count all those barcodes and those unique identifiers, and then they rank order them, and they are able to generate an entire continuum of data, which is absolutely critical for models these days. Because if you're just doing an experiment and you're just looking for those winners, that's like trying to learn a language with only reading one page in a book. It's just, it's very much incomplete. But what you actually need is that full gradation of data, losers, winners, and everything in between, so that way the model can understand that full context and then take that and make accurate predictions for the next round of experiments.

And you said models is like AI biotechs, and give me an example, like, are they customer-facing or do they just deploy products, right? Like give me an idea here.

So there are a lot of, you know, free models out there as well as a lot of companies that have their own proprietary models. For example, one of the hottest things these days is antibody discovery. You know, antibody drugs are, you know, tens of billions of dollars of market there. And so, you know, people are taking these libraries and running them against, they're usually trying to bind against like a cancer target or things like that. And they're trying to find, you know, not just the winners, but what are the things that influence how well that this antibody binds to the target? 'Cause they wanna make a drug. They really want something that is just gonna work. And, but it's, you can't just look at one thing. You have to look at the stability. You have to look at the manufacturability. You have to look at all these sorts of things. And so people are running experiments, not just in one dimension, but many, many, many dimensions to be able to understand really how these proteins, you know, interact or form these complexes. Because just taking the simple approach, we've done that. We've done that for decades now. You know, we have Keytruda. It's an amazing drug, things like that, but it only gets you so far. And so, you know, people are constantly refining these AI and biology models to be able to make these like zero-shot drugs right off the bat. And what that really translates to the patient is lower drug costs.

You're muted. Oh, uh, yeah, you guys scaled the company to impressive feats already, like in a year, right? Like almost going for 2. Uh, how do you scale that company to the next level? What are your main objectives for this year?

So the real thing there is growing our client base and establishing new products. We are constantly bringing new products to market to be able to expand our portfolio and meet customer needs more quickly and economically than ever. We are developing some new products that, like I said, will meet these customer needs but don't require like a large expansion of headcount. You know, that's one thing. We want to scale our company, but we don't want to just like become a massive company, if that makes sense.

What's the definition of not massive, you know, and why not scale in a very competitive market?

Yeah, so the complexities that come with that, you know, we saw that at Ginkgo. You know, we've done that previously at the previous company that was acquired by Ginkgo, Circularis, as well, is when you, when you're expanding a lot of these areas, you can dilute out like your core offerings, and that can kind of, you know, compromise your ability to do what you do best. And so we are being, we are growing, but we are being hyper-selective in the areas that we grow so that way we don't degrade our core product.

In terms of revenue and employees, like, what's— what do you mean by limiting the scaling?

So, um, you know, we're not trying to grow to like a 50-person or 100-person company. You know, the, the amount of, uh, infrastructure and overhead that is required for that is, uh, can be crushing. And we, we've seen that in the past. And, uh, so we are being very, very judicious about how we grow and mostly just being able to maximize our revenue, the amount of revenue that we can capture at each stage. And with the main thing, we just, we really don't want to get over our skis.

What's been challenging running that type of company?

It's startup land, and so a lot of it is sweat equity and putting a lot of time and hours into it. And the main thing right now is that, you know, with anything in life or capitalism, timing is everything. If we had started this company a year ago, like, we would already have had to close the company because the people costs to do what we would have had to do a year ago would have made— we wouldn't have been able to keep up with it. But because of AI and everything that's happened in this last year, the last 6 months seem like 6 years to me. And what that's done is allowed us to be able to grow as a virtual company much more than a human company. And so we are a 5 FTE headcount here at Flock, but I have 15 virtual employees that I've coded and built that run a huge amount of the company in the background. And I have a real strong feeling that this kind of like 3 to 1 or even like 5 to 1 or greater of virtual to human employee is really going to start to become the de facto standard very quickly.

The team itself, like, who do you hire versus who do you— what do you outsource to AIs nowadays?

So I would say that the whole thing with FlockBio is getting the band back together. Like, everyone that I work with, I've worked with for years. And so there are a lot of folks out there that are looking, you know, for jobs and things these days, including my my friends. And so, I kind of hired, you know, we kind of hired our friends. And what that does is allows us to really accelerate what we can do right off the bat. You know, we're not really spending a lot of time onboarding folks or training them or teaching them all this sort of stuff. We already know what we're doing. And then what fills the gap on the virtual side is like, you know, none of us, we're scientists, none of us are accountants. So, you know, I have, you know, a Flock CFO agent. I have a Flock go-to-market agent. I have an EH&S agent, so on and so forth, that are able to help us, you know, in these areas, you know, fill in these gaps in expertise. But you can't just let them, you know, have carte blanche or whatever. It's very much a lot of trust but verify. But what it does is it just really increases the amount of things that I can parallelize all at once.

Well, Matt, thanks for coming on. Where can people find out more about you and your company?

Yeah, so feel free to check out our website, flockbio.com. You can send us an email at info@flockbio, and we'd be happy to chat about folks' experiments, upcoming projects, things like that. Or even if you just have like general questions about pooled libraries or how these things could work, we're always happy to lend our expertise and have a chat.