LA Venture Podcast


Peter Lee — Embark Ventures

November 4, 2019


Peter is a founder and managing partner at Embark Ventures, one of the very few deep tech/frontier tech focused investors in town.  They write $500- $1M checks into pre-product companies with important science behind them.  

Peter is full of interesting thoughts on robotics, brain computer interaction and tech transfer out of universities.

Full transcript

Hey we're here at embark ventures with Peter Lee. Hi Peter. Hi. How are you guys. Thanks for doing this. So embark is a seed stage deep tech frontier tech fund here in Santa Monica. Across the street from you right here in the Bermuda Triangle all the V.C. funds upfront bonfire I across the street I run into fusion. Yes. So anyways it's fun to be in your beautiful office and to get a chance to talk to you. Great. I love to be here.

Great to see you Peter. You too. You guys know each other well. So I'm probably interested in hearing more about your background because I think you guys probably crossed paths more over the years. Yeah.

So I've been in L.A. V.C. now for 12 plus years. But it might be helpful to hear my story just because it delves into what our fund does so early on my dad was a professor pasty and I actually thought I was gonna go down that route. So I went to M.I.T. did my undergrad MASTERS My thesis was in robotics and I summer at Los Alamos National Lab doing robotics. So I was heavily down the research path. What did your dad professor of nuclear chemistry. Cool yeah. And so I applied for a PHC at M.I.T. got in but this was in the late 90s when the Internet boom was going crazy.

 And so I was like well maybe I'll work for a couple of years first before diving right. And so I deferred my paycheck he ended up joining a startup second employee still working out of the founders dorm room no funding and after 3 1/2 years we'd raised 50 million dollars in capital grew to over one hundred and twenty people and had the most amazing experience in a startup. So I was doing a lot of ultrasound and optics work hardware design went to China a couple of times that of manufacturing and got to deliver product from a PowerPoint stage through selling billions of dollars worth through Best Buy you know OfficeMax Staples and the opposite of the PHC experience.

 Yeah very very different. After about three and a half years my dad mom we're sort of bugging me because they're both scientists you know restart the party and so I ended up going back for the Peachtree lasted about four months and then realized I just could not finish it. It was just I think the sort of experience was maybe depending on what perspective the best thing or the worst thing I ever did.

But the fast paced environment was amazing the amount of learning and authority and responsibility were given at the age that I was was sort of vastly beyond what I had gotten a large company and so I end up dropping out of the party. I decided to business school instead just because I had a lot more interest in not just the end during development but also learning product management and general strategy. So that sent me down a different path. Somebody at Microsoft wanted McKinsey for a couple of years really miss a startup experience so I went after McKinsey and left to join another startup and then after that company was acquired I got into venture capital and was that Baroda ventures.

 No no. So I actually started at a firm called clear some venture partners which has morphed in some ways to march capital. So to the partners from clear Stone left found march but back in the day there were half a dozen venture funds I think none of which really exists now at least in the same name or form. So there was clear stone and Palomar and zone ventures and GAAP GAAP which became upfront front yes. Now part of the one that's the most consistent. But in terms of the the partner profile then versus now it's almost completely changed over except for Eve I believe.

I think that's right.

So what made you decide to come to L.A. of all places to become a venture capitalist back then.

Yeah it was a wife thing. So I met my wife in Boston. She was from L.A.. And as long as I was in some big town I told her I can go anywhere. And so she grew up in L.A. so we just had to make L.A. our home.

And when did you move over to Baroda because when I met you that's where you were. You're the lead and pretty much only GP there right.

Right. So the Britta story I'd been in L.A. doing venture for a couple of years and I met David Burnett who is a very prolific angel investor. His most famous for founding geo cities. So back in the late 90s Internet boom he found geo cities took a public and then Yahoo acquired the company for something like four billion dollars. So he became an angel investor. I was doing a lot of investments in the L.A. local scene. He'd built geo cities here I went to USC. And so when I met him we were just chatting about his own involvement and venture and what I was doing.

 And after a couple of lunches we were brainstorming about where the opportunities lie. And he invited me to come and run the seed fund for him. And so this would be in 2009 when I joined Baroda and Ellen were there eight years seven seven eight years. So it's a pretty long stint. Yeah yeah.

And when we met. When you're doing Brody you weren't particularly focused on hard tech now.

Yeah. So so you know if you rewind back to when I was doing venture in L.A. through a last couple of years the venture scene L.A. was very very small at the time and it was very hard for a fund based in L.A. to be too vertically focused. You had to do pretty much everything be a generalist if you wanted to make the vast majority of investments in L.A.. And so we did consumer web and sass and tech and e-commerce and ad tech. So anything under the sun. Fast forward to today the L.A. ecosystem has grown tremendously.

 And when I met my partner Adam Barak a pain. We both felt like there was an opportunity that was unmet in L.A. where there's a lot of funds a lot of growth a lot of great startups. But in terms of the hard tech areas so areas like robotics manufacturing materials science there was almost no capital especially early stage that was willing to make investments especially pre product in L.A. and where those founders coming from down here were where.

Where do you what's the genesis of these companies that are doing more frontier tech.

Yeah. This is the other place where we felt like the option was it wasn't just that they weren't funders but if you think about Southern California broadly you have amazing research institutions from Caltech all the UC is UC Santa Barbara UCLA and UC San Diego and then a lot of the aerospace and defense industry is down here. So you have the north through ups and Raytheon's and then in San Diego you have Qualcomm and then more recently obviously have Space X and so there's a lot of very strong technical talent but less entrepreneurs who are spinning companies doing hard tech.

 And part of that we feel like was because there was a lack of capital. And so when we started to talk to entrepreneurs in the area who were interested in doing heart tech companies what we consistently heard was We have to pitch the Bay Area because those are the only investors who are willing to write checks and generally especially if you're early. The response was once you move up to the Bay Area and we'll consider it or come back to us in the next round. And so we just felt like that capital source for people who want to stay in Southern California was not available.

There's a lot of bio in San Diego. Did you write to do that at all.

Initially when we started a fund we stayed away from it. More recently my partner paying has been spending a little bit more time in bio so we won't do peer therapeutics like new drug discovery. But I think our interest lies more in applying some of the A.I. machine learning to other industries specifically bio. And so we're looking at some med devices but also business models where you can advanced like synthetic biology and biology based on some of the new technologies which I think is kind of good.

Just give us the basics of embarking again. So your seed and proceed. Right. So many people have products some maybe don't have products but they've done the research. Tell me where you come in.

Yeah. So the profile of embark with a 45 million dollar fund. We invest again precede and see it as you mentioned our distinction. You know maybe differs from other people. Almost everything we do is pretty product launch. And so I think out of the 15 companies we've invested in 14 we're pretty product. And so we're always early. Our distinction is when we do a seed a typical seed investment for us. Generally there is a very strong technical founder as well as a business founder. And those would be cto ceo with probably a significant amount of technology development already gone into the company and they're looking to get to first product.

 And so the distinction for us is we don't invest in scientific risk. And so if there's pure science there's just you're still trying to figure out that's better for NSF in the laboratory and a university. But once the core science has been figured out and now you're translating that to building a product applying engineering to actually deliver a product we would like to enter at that stage.

So most of the companies on our round of financing which is generally the first institutional round we want them to get the first product first customer oftentimes the difference between science or the first step between science and a product is the ability to scale what's happened in the lab to something that's production skill. We take that risk.

Yeah I mean I would actually say that that is part of the science differently but in science and engineering. And so if they're still trying to understand the phenomenon or how to make whatever they're building replicable I think they still need to do more science homework if you will. Once they have a repeatable process they understand the science behind it then taking that to manufacturing and scale it becomes a risk that we're willing to take.

I'm very interested in the science interface if you will because I don't think most people that we've been talking to have as much interface with the university system or with NSF of the world. Can you tell me a little bit about non dilutive funding if people are getting SBIR or what you run up against there.

Yeah so a lot of the companies especially coming out university will raise six sigma of an amount of non capital. And so we we love that just because it was more capital. Yeah that sounds like us. Obviously they're very focused on what they can do with it and so you have to make sure it aligns with what the company is doing. But I think it's a great way to extend runway especially when there's a lot of questions.

How significant when you talk about significant candidate of funding how significant and significant is it. I run up against as the IRS it's the only sort of place that is that the most common is that one of the more common.

Yeah. Oftentimes it comes from you know NSF grants and generally you're looking at you know quarter million and first phase growing to a million plus in later phases really. And so oftentimes you can find companies that raise one to two plus million dollars and on the list before we even either enter or alongside our financing.


How about as some people know I live across the street from Caltech I'm always interested in how about these sort of incubation type programs. Am I. Be Yeah. And do you recommend those that do look at companies coming out those programs right.

And this delves into what our version of precedes so I'll address that and then get into the core program. So for us our version of precede because it's still early still pretty product is oftentimes we'll find a technical only founder and they have a technology they want to build a company off of it. But maybe they're not so far along to have recruited an experienced CEO or business co-founder. And so we see an opportunity to invest a small amount of capital to get them to maybe a prototype demo stage where they can then bring on maybe a business CEO or business co-founder.

 And so that's our version of precede it's a little bit earlier but we see enough of those opportunities that we feel like it's an opportunity in terms of the NSF core. The big challenge of dealing with scientists engineers is most of them spent their entire life in the lab and they've never built a product to you know for customers and commercialize anything so. But they know their technology very very well. And so the NSF I call what we like about it is I believe part of the program is they force you to have like 100 or 200 customer meetings and it get scientists who have this idea of what their product or technology is to go out and talk to customers and they actually learn what is critical of versus is not critical because scientists they love their technology even though they don't necessarily know necessarily how best to go to market.

 They might look for a small opportunity and it ends up being a market that just isn't interesting from a venture sale perspective. And then oftentimes they are so enamored with the technology that the market only needs a certain amount of sort of advancement and maybe they can do 10 times better than that. But if the market doesn't need it it's sort of useless. And so I think it trains them to really tailor what their technology advantages to what the market opportunity is. And by going through these 100 200 customer interviews they come out on the other end much more aware of how to take it to a market.

You talk about matching technical founders with it with the with a non-technical not necessarily on technical but with a business person. How do you do that. Do you have to have a stable of people who are waiting for four great tech to to get behind and how do you make sure that that matches good it's hard to do any kind of shotgun wedding.

Yeah so. So I would be upfront. Are pre seed investments are still pretty early. We've done two going on three of them out of our fund. So we're still experimenting a lot of it is I. We don't initially have a whole stable of executives who are ready to go but a lot of it is just sort of organic as we get to know them. A lot of our diligence is with people in industry who can help vet the technology but also that the market size. And through those conversations sometimes we'll run across executives who may be interested in jumping out of corporate into a startup.

 And so we don't necessary have a very strict game plan of exactly how it's going to work. But it's a little bit organic and oftentimes we just fall in love with a founder who has something interesting to show signs of being a great entrepreneur but might need a little bit of guidance and help on building a team.

How important is IP when you're looking at this.

Yeah I mean IP isn't patents. I think it's it's it's helpful but it's not sufficient in a way. So. So when we get into a company we definitely like companies that have more defensible proprietary technology at its core. And our definition of it because you know does that mean they have a patent or not. It's more of we use a lens of if there is a smart team of engineers who generally know the market you're in could they go in after hearing what you're doing or seeing a demo and replicated in under a year.

 And our view is if they can do that it's probably not defensible enough. So for us IP and defense ability is more around how easy is it to replicate and how much competition there will be for us.

We're less set on having to have patents. I think it's more important later on down the line when you have a lot of corporations really like to have a patent portfolio as assurance that there's there's more defensively there. But for us it really is more round.

Do they have a headstart versus competition so that makes a lot of sense that potential acquirer might pay more for IP.

So do you work with the tech transfer offices and not just around the IP. But if I'm spinning out of the lab do I still get access to the university equipment or is that part of what the tech transfer does.

Yes so we've had some dealings with tech transfer. A lot of times the more reactive to what the entrepreneur wants and so if an engineer really wants to spin something out hopefully you have more sophisticated tech transfer offices so obviously you know Stanford M.I.T. are really really sophisticated. They know how to work with companies and they do a great job. Some of the other universities I think are are working their way to get better but they are responsive to the entrepreneur and trying to figure out how to make it easy for them to spin technology out in terms of using lab space that is actually a separate discussion.

 So oftentimes they can get funding or can rent lab space. And so if they can do that again it helps extend runway because their capital needs are a little bit less right now.

I mean if we could shift gears and actually talk about some of the substance of things I find the process interesting but maybe the substance of what you're looking at. I know you like robots right.

I was I was thinking that to be great to tell us about an example company that's that's here in L.A. that know that fits these criteria and invested in so about half our portfolios L.A. half of it's outside one of the first robotic investments we made is it comes about India robotics.

They do warehouse automation and so this is when you order something off an e-commerce site and somehow get says some guy in a warehouse who grabs all reform walks around a warehouse and grabs things off the shelf is very constrained environment very orderly hopefully warehouse. And so having autonomous robots that can do the work for this person makes a lot of sense. And so that one was a unique case because the founder with this was I think his fourth startup. All three of his previous starts were successful his last one cast sold for I think 350 million dollars of horizon.

 So he already knew how to start a company very sophisticated there. He also had two co-founders who came out of a PHC lab out of USC doing robotics and so very strong team technically but the CEO who is very sophisticated at bringing technology to market.

But I think that you've also given the example. I know you spoke at innovate Pasadena. I know you've given the example of if it's just a off the shelf robotic arm and off the shelf can be a computer vision. So this was something more than just sort of generic warehouse automation.

Yeah. So for us when it comes to defense ability one of things we'll look at is there's a lot of companies doing some kind of automation task where they'll take an off the shelf robot arm. There's a lot of computer vision out there and they just do some task and say hey we're going to replace a human labor with this. For us it's not as interesting if that's the only thing they have. Just because we look at it as you could take any grad lab grad student lab in a robotics university and they could probably replicate that pretty quickly.

 And so we're looking for either something on the hardware side that's more unique or on a system level that's more defensible. And so for India I think there is more of a system level defense ability around how they integrate with a warehouse and some of the algorithms and optimization about how they move goods around so what are some of the interesting problems in robotics that remain unsolved that you you're looking for.

Is it is it more on the hardware side or is it more on the control side.

Yeah I think I think there's a lot of the avenues are all open for for improvement. So on the hardware side there's a lot of generic robot arms that can do basic tasks but I think we're at a place in robotics where we believe a lot of the great solutions that will really provide the right level of consistency and sort of ninety nine point nine nine percent of the time we'll succeed really need to be custom hardware to do it well. Also the robot arms are still very expensive though the prices are dropping.

 And so I think there are certain cases where robot arms makes sense where you need more flexibility but oftentimes you need custom hardware built specifically for one type of task. What does a robot arm cost. I mean some of the larger ones. One hundred thousand plus I think now that are dropping to you know 20 30 40 thousand dollars for a slot smaller arms. But you know if you're if you're going into an operation where you have a laborers being paid 10 15 dollars an hour to spend one hundred thousand dollars on a robot arm to replace that person the hour why doesn't work oftentimes.

 And so if you can now provide that arm for 10 15 20 thousand dollars the ROV becomes a little bit more interesting. There are some startups one here in L.A. elementary robots that is I think trying to come out with a two three thousand dollar robot arm much much lower cost.

And if you can actually do a task using that arm the ROV becomes very very apparent very quickly but you're pretty interested in not like androids right but you're just like manufacturing right.

Heavy industrial sort of use cases.

Yeah. So for us on the robotic side we are more ROIC driven. And so oftentimes when you have industrial use cases you actually do a calculation on how much are or why the customer and company can have when you get into consumer robotics and the value prop is entertainment or enjoyment it gets a lot harder for us to really evaluate whether that's going to be a good investment long term. This might be a silly question but what's the difference between a robot and a machine. So yeah yeah it's kind of like a robot my clothes my dish cleaning really.

 So I think there is a level of intelligence and adaptability that you look for. And so you have a lot of even like for example a car manufacturing plant you have massive use of robots but generally they are dumb robots. They are programmed to move from this coordinate to that coordinate to pick something up and move it somewhere else but they don't learn they don't adapt. And so if anything changes in the assembly line it breaks down because they don't adjust to the environment. And so for us that is more of an automation task versus autonomy.

 And so we're more interested in companies where the environment might be a little bit more chaotic and there is intelligence behind the computer vision as well as the robotics to adjust to the environment. So one example of that is a company called Safe A.I. that we invested in that's building an autonomous driving platform for the heavy equipment industry. So this is mining trucks large construction equipment and so you know one use case is called the load haul dump cycle where a truck will go to one area of the mining site get loaded up with a bunch of dirt or whatever haul it off to another area and dump it and they will repeat that loop all day long.

And so it's a fairly automated task they're doing but it's on a mining field and things will change whether lighting conditions and so being able to automate that task one reduces some very expensive labor allows you to run overnight and when you can run 24/7 our life becomes more apparent. But the environment is still chaotic enough that it would be hard not to have some good machine learning intelligence behind what they're doing. And one of the other things we had around automation autonomy was I'm super excited about autonomous driving and having cars on the streets.

 The problem with that is there's so many edge cases that you have to deal with. You know traffic light goes out and you have a traffic cop in the middle of the lane. How do you respond to that. And when there are a million different exceptions that can happen it's very very difficult at this stage where adding autonomy to solve that which is why I think you're not seeing as as fast a progression of autonomy on public streets. And so for us we are thesis was to go into more strained environments.

 And this is a mining field this is a warehouse. We have another company that's entering the construction industry. And so the environments they're operating in are much more constrained. You don't have kids running cross the mining field or ball's rolling across the street. And so the number of edge cases is reduced a lot. And I think autonomy right now is capable of satisfying those conditions. But you know is it gonna be five years or 20 years. I think there's still a lot of people who have questions on when full sort of public street autonomy is really going to be there because if anyone has ever lived or spent time in Boston trying to have an autonomous car in Boston the winter operate well is is not going to be an easy thing.

The drivers are so bad that you cannot apologize to any Boston listeners. One funny example I remember hearing about an autonomous car dealing with us and is something we deal with every day. They tried to enter a highway on an on ramp and highways totally backed up and that Thomas vehicle because it knows not to crash the car just sat on the on ramp and it never moved. But for any human it's a fairly straightforward test. You know if it's busy you can't just sit there and say these sort of play a game of chicken where they keep edging forward until someone gives in and they stop and you're able to get in.

 It's very easy task for human to do but to program an autonomous car to push yourself forward until someone gives up and not want to hit you is not easy to do. And so a fairly straightforward task that you experience every day especially in L.A. I think autonomous vehicles will won't be able to solve that for a while.

Yeah anything that involves communication between the driver or robotic driver and other drivers or pedestrians is really all right.

Right and safety is always the biggest concern for all these platforms and so if you can't address that well with random human behavior it's going to be hard to get to a point where you can deploy that.

But another perhaps related area that I know you're interested in is the brain computer interface right. Yeah. And I guess is some of this sort of the big picture sort of an interesting robots and brain computer interfaces. They're like reverse engineering humans to some right. Right. It's a pretty fascinating picture subject.

Yes so I haven't made the investments yet the BCI space but it's obviously Fascinating I'm sure people saw the Elon Musk announcement about hooking up a chip to your brain and you know for us you know I think it's an area that we think it's fascinating to be able to get you know robotics and automation and now you have human either interface or control with robotics or other systems. We still are trying to form our thesis around you know what the right investment is gonna be especially for a venture fund timescale. But there's a couple of ideas around sleep enhancement lucid dreaming that we're exploring that we're actually think about potentially incubating but we're well we'll see where that leads us.

So I actually didn't follow that. I mean I saw the announcement that what is BCI.

I mean there's neurons and electrons. Can you break it down.

Yes so I think the easiest thing is there's been a lot of technology to read brainwaves that's been around forever and so you'll hook up some device to your brain read the brainwaves and you can see signals coming out of it. You're able to potentially instead see their thoughts in terms of what the intentions are and so you might have seen like a quadriplegic play hooked up to some interface where they're able to move a robot arm. And so that is the first stage of Bci. What's become more interesting to me is not just reading the brainwaves but then actually can you influence something back to the brain.

 And so using things like low low electrical impulses low power ultrasound focused ultrasound. Now you can not only read the brainwaves but you can actually potentially affect what's happening in the brain. And so that's sort of a new frontier that people are exploring and would that be for like physical feedback or more nefarious things. Well I mean I guess it could be for a lot of stuff so currently some of the early applications have been more for psychology and so they found people to have like deep depression if they use I think low electoral stimulation you can actually affect certain areas of the brain to reduce some of this depression.

Some of the things that we've been looking at are I'm very interested in the whole area of sleep study and what you can do to hack sleep and you know you hear about people who can sleep for four hours a night and be totally well rested.

Hey hey you know I hadn't by eight hours and I'm still tired.

And so you know if you look into sleep cycles you actually find that there are different stages sleep once physically area called Deep Sleep which is where most of the rejuvenation comes from. And so we've been looking at ideas around Will can you cook up something your brain detect when you're in already embassies light sleepers deep sleep and then provide some input back into the brain to try to extend the length of deep sleep. And so the idea of you know could you actually have a device that someone to wear detectors the patterns which can be done now but then influence the brain and actually adjust you know that the sleep pattern within your own head so that you can actually get more efficient sleep and sleep less.

 So things like that are interesting to us but it's very new still.

So maybe we just touch a little bit more broadly on embark and there's you and there's e-mail and I don't know exactly what he's interested in but yeah I guess so.

So cool. Yes. So he's doing more of the sciences so he's more involved in material science as well as some biotech areas whereas I'm doing more robotics and manufacturing. But in general were early stage seed investors typically writing half know into a million dollar checks investing as much as we can in Southern California but also outside the country or sorry outside the region in the US.

And how should a founder think about approaching embark.

Yeah. What's the best way to get to you. Yeah. I mean I think it's priced similar to most venture funds where if it's just coming over the transom it's it's a little harder to get our attention versus if we get into rowed by people we know and so. But in general you know we're seeing because we're so focused on deep tech. There are less volume of companies that come to us and we can spend more time and diligence and so we're spending time at universities we're spending time trying to find alumni coming out of Space X we're looking at some of the larger companies that do research development in Southern California.

 So I guess we maybe fish in different ponds than a lot of the other species do. But it's still the same process in terms of finding good founders.

But I imagine that a lot of academics don't have connections in D.C. right. Right. How should they go about funding.

I mean we're actually after spending a lot of time on university campuses and so hopefully through tech transfer we've built relationships with different labs at different universities we've built relationships with and we're holding events also where we invite people from universities grad students to come in and learn more about venture so we're still early from as a fun perspective but hopefully we found you know specific areas within universities to get access.

Who else is in the L.A. ecosystem.

E Do you run across and that could be other venture funds or just other people in your broader space that aren't just like the university tech transfer office in in L.A. There's not very many funds who will invest in hardware as early as we do.

One Fund riot ventures is one that we've co-invest with. They'll do some hardware. There's a couple of the funds who will dabble in hardware but I would say not necessary to the level of tech that we do. And so generally we're actually looking outside of the L.A. area to fund co investors or syndicates just because in general L.A. is still an ecosystem growing well from a financing perspective but from a deep tech perspective still very very light got it.

There's a few of the sort of more make your make it right and make it right. Yeah.

So they have a small fun but they also have more of an incubator model and so they will bring companies in that do hardware. I think the difference between us and them is they will only focus on hardware for us. Generally the companies we invest in and have some hardware component but our interest is more on around the defense ability and so they will do some companies that is hardware but not defensible at all. We won't touch that but our interest is more finding the sort of hard core defensible technology as our base is doing like a one second time time.

I forgot to start my time 33 40. I think there's a lot of good stuff. I find it really interesting. Yeah anything you want to talk about that we haven't gotten to.

Yeah. This is really like. I mean I don't know it like I do my thing too like no. I mean one of the question I was very generic sort of about.

You've been at Veasey now and a few different places I've been in B.C. for about like four months just working with David like there's a little bit of a question there around his site. I know the question is well you don't advertise your newbie ism fundraising man you know I just I always like to ask about advice to founders. I say it is a positive. All right good. All right. You talk about people just screwing up their capitals that of thing. Yeah that's the question.

Yeah I ask that question and advice you maybe advice for especially more technical base founders.


So I'd love to also open this up to our two hour entrepreneur founders and just ask you advice for those folks.

Yes so oftentimes because we're investing in hard tech companies we do meet grad students people who were purely on technical side and generally they just have less knowledge about startups. And so I'd say a couple of the major mistakes we see are the obvious one is you know technologists falling love with technology and not really knowing how it applies to the market. And know they talk about a product market fit and founder market fit. There is some element of sort of technology market fit. You know you might have a technology that is you know two three orders magnitude better than anything out there.

 But if the market doesn't need that and they only need something that's you know 1 percent better then you sort of overdeveloped and maybe you know offering a product that isn't really as required. And so I think there's an important element of where does your technology advantage really match up with a need that no one else can solve except you. And if you can find that and it's a big enough market then you have something interesting. The other piece of this is not necessary Just technical founders but any first time founder you know you have people who they they don't know what the general process for fundraising is the earnest and cap tables and they might raise money from an unsophisticated Angel and end up doing things at the cap table that you know might not be advantageous long term.

 And so we just hopefully will you know guide early founders to find advisors who've been there before because oftentimes these deep tech companies just have to raise more capital especially if there's hardware involved it's just more expensive than than a typical software company early on.


And then Peter it's been great to have you as you said a good opportunity to catch up and hope to see more of you. Yeah sounds good. Thanks so much for this. Thanks Peter. All right. I was going to ask you about this.

This capital efficiency thing because it's one of the reasons we've shied away from we've walked away from a few deals that looked really really interesting and that have hardware is like oh gosh you're gonna need a lot of money. I'm not gonna be able to keep up. Yeah yeah. Do you think what we do.

I mean at some I think the hope is that if we are investing in something that has that big of a technical advantage maybe the outcome could be bigger. And so even though we get diluted over time maybe the value of our holdings will still be big. But in general it's going to require more capital. So a typical seed raise in deep tech has actually price running two to four million dollars whereas oftentimes you'll see that being you know a lot less requirement on the like up your software SAS company.

So that that exists. But yeah I mean we're still learning this as we go.

Did you look at the pizza robot out of space x. Yes I look very closely. Yeah we were we were we were actually interested but it it's become a very competitive space. I was there to do anything besides pizza robots and I'm going to. Right.


And in know software coming back to the sort of capital efficiency thing. Maybe Jason again a software company today teams great. They built a battery for the dragon. I'm just gonna invest in it right. Right and pray for the pivot. Yeah but with hardware it's harder to pivot. Yeah. It's just more expensive.

Yeah I mean a lot of these companies will build some hardware that can be reused and so there's a lot of stuff in food automation and so maybe they're making something for smoothies but then it can be applied to bowls and salads and things like that. So there is you can stay maybe roughly in the same genre and extend your technology a little bit but you are a little bit more locked in. Yeah. Is it pizza robot. Yeah. Yeah I did. I did. But it's like a 80 billion dollar delivery market for pizza.

Yeah and it's really just a can't. Yeah. Benson. Benson.

Yeah. I just kind of want to move the state of the world forward and having pizza delivered pizza doesn't do it for me. You're right. I gotta say I got a you couldn't I just for me. Motivate me. Maybe you. Any ideas.

I know I did but for us on the food tech space the interesting parts are we we've looked at a number of companies and a lot of times they are. Here's a robot arm or system that we sell and touch politely or some others other establishment and the the RSI benefit is you get rid of Labor but that's actually not that big because you're not saying that there's not expensive labor.

And so the other piece that we look at is real estate which is very expensive for a lot of foodservice companies. And so we've been more looking at mobile food kiosk based opportunities where if you can automate the making of whatever products you get rid of Labor as well as real estate and those then you start getting to our why that becomes much more compelling.

And right now the labor market is super super tight. So if you're there then it might be in right. Yeah in a worse economy in the worst economy.

Yeah but even that I mean like replacing a 15 dollar an hour laborer that's not that skilled. I think it's really hard to provide a couple hundred thousand dollar solution and say you replace two people on the line and make a.