Nominal raises $80M at $1B valuation to bring hardware testing out of the Excel era

Then let me also tell you about Reream. One live stream, 30 plus destinations. If you want to multiream, go to reream.com. And our next guest is here live in person with us in the TBP and Ultradom. We have Cameron Record from Nominal. Welcome back to the show. Good to see you.

How should we kick this off? Can we just hit the gong right away? What happened? Tell us. Tell us what happened

today. We are announcing a uh $80 million raise at a billion billion valuation.

That was a loud one.

He broke the mallet.

He broke the mouth. That's a powerful powerful omen right there.

I'm excited.

We're announcing our latest uh our latest financing led by Founders Fund.

Fantastic. Fantastic.

And what do we what do we have here?

Had you been working with them before?

We had been working with them before. Okay. Yeah. So they founders fund have been an investor um since the seed round.

Okay.

Um

through Delian Trey both.

Delian and Trey. Yeah. Both. Um and really uh I think really how this

Yeah.

came about was um I think Founders has a really good perspective obviously on the evolving hardware landscape broadly speaking.

Yep.

Um but a unique perspective into particularly Androl.

Sure.

We announced our partnership with them.

Yep.

Uh two weeks ago now. know if it's working or not.

Exactly. They're going to know if it's working or not. And so I think um you know some direct feedback from them and and other uh FF portfolio companies. Um and so yeah, we we were not raising I mean we were on the show

uh eight months ago announcing the

the series B led by Sequoia. Um

but we got approached by Founders Fund uh at the end of the year really over the holidays um with a pretty good offer to lead a like preemptive financing and it made a ton of sense. So

that's great. You were like why not make a magazine?

Yes. So I brought

the magazine first sort of reintroduce the product for those who didn't see the previous uh

Yeah. So nominal we build a uh a platform for hardware test and operations. Right. So we're managing everything from um the end of the manufacturing process. Uh so then quality testing, inspection, end of line through to more like lab testing, benchtop testing. Okay. think power supplies, DAXs, oscilloscopes, instrumented hardware.

And making a submarine,

they need to test the battery that goes in that submarine. As soon as it comes off the manufacturing line, even if they buy it from a supplier that they trust, it comes in, they have to know that it has the right voltage and

and that's exactly honestly that is a really really like we're seeing that use case really explode in the product. Um I think particularly as there's been a big rush of more onshoring manufacturing. Frankly, I think quality in the entire industry is sort of going down as there's a little compression effect. So, um people making yeah components, widgets, uh you know, servos, motors, uh etc. Um when these uh the OEMs, the bigger companies, the Andals of the world receive them, they they need to run through some yeah some initial.

Does your business grow as businesses that you work with shift from R&D and testing and prototyping to actual scaled manufacturing? Yes. Yeah. And that's a huge part of the thesis like honestly like some of the earliest

like what you unlock.

Yeah. Exactly. Some of the earliest vision of nominal is like testing is a really powerful place to start because every organization is kind of always testing if you're iterating. There's normally some budget. There's a little bit of risk tolerance to try new software, new tools, new processes, get whatever advantage you can squeeze out. And then a lot of the thesis is like if that works, you want to use the same software uh when you're actually doing highcale production manufacturing. And you want to use the same software again when you deploy your asset. Let's pick a submarine, you know, um, wherever it is in the world. Um, submarine is probably a bad example. Again, I was a former submarine officer, so cool.

Know a thing or two. Um, you're not going to get a ton of live telemetry coming off of a submarine ideally. But, um, but if you deploy an asset in the field, um, you want to be able to link all of that telemetry sensor data logs back to the the system.

Interesting. And and and is that for uh like testing like the feedback loop for manufacturing? So it's like let's switch over to like the sensor tower analogy. You put up a bunch of sensor towers somewhere. There's a bunch of telemetry that's coming off that. Not just the purpose of that sentry tower, which might be uh object identification, but also the battery.

The battery is a really good example and we'll we'll use uh you know uh sentry tower counter US tower as an as an example often deployed in places in the world that are very hot. Yep. And so um there can have a lot of battery issues. It happens all the time and you want to be able to correlate if you're getting anomalous or or spurious readings on the battery. You would use something like nominal to be able to monitor that. So think like how data dog is monitoring you know

y

like a a server or uh or you know generic software like we'd be you set up a conditional to monitor any weird anomalous behavior in that battery. And then you want to be able to link that correlate that with the physical assets maybe the the group of batteries that you receive from a supplier. Um, and you want to be able to trace that backwards through your whole process and say, "Hey, actually those batteries that might have similar issues are deployed in these like five areas because I can track those assets and understand and sort of, you know, think think that way

because the battery might perform differently in different temperature conditions, even just how much the CPU is turning

really. It's just like Yeah. And I think what nominal is unlocking is really just like the ability to do this correlation at scale. And frankly like what we we the way we we store and organize all of this you know massive amounts of telemetry and sensor data um is just like a much uh you know more thoughtful approach that you can query at scale and sort of make these types of connections really uh the industry status quo which is a good segue into um the the stuff we brought here. I think the industry status quo is like it's just there's a lot of legacy tools in the space. So it really is um it is Excel, it is mat lab, it's PDF, you know, it's it's old old pieces of uh of software that have not really kept pace with the

Yeah. Last time last time we were at the track getting some laps in the the software that that the team was

using to track lap times and you know uh just all the different data coming off the car.

Uh

what literally look like '9s era software. Mhm.

Talk about the partnership with Pratt Miller. How's that evolving? Are you are you going to add more teams? Is that like a great feedback loop for maybe more traditional

I mean so so last time we uh Bryce dialed in? Um is you know this is very cool being in person, but he was at the the pits and uh that was pretty sweet. No, the partnership is going great. I think um I'm personally very excited like we we um we get to go and attend a bunch of these races um and we get to bring uh our customers um or potential customers which is great because they get to see nominal in action um and so it's a really like nice like way to kind of do that. Um then yeah the partnership is uh is expanding. There's

I won't say too much on that but I think we're um we're expanding in the automotive and racing sector kind of broadly um based on the success there. So yeah,

very cool.

But yeah, I uh

through this

Yes. tools

for progress.

Tools for progress. Yeah. Our sort of slogan for this financing has been um ambition is timeless, tools are not.

Uh and so uh going in the history lesson, you know, it used to be in the 80s 90s uh you know software was

uh industrial software was serious. It was purpose-built for you know aerospace, defense, mechanical engineering, etc. We then entered two decades of very unserious software.com you know click optimization uh pass through you know uh metrics etc. Um and then nominal is trying to get back to serious software for the next five decades of engineering. And so what we put together here a very tactile you know thing. Um this is this is waterproof weatherproof also um which it would have to be. Um so tools tools for progress. Um this is an old school catalog. This is where like the best tools in the world uh and the best companies in the world used to

uh write about their products in paper.

It's kind of crazy to think, print it up, send it to engineers all over the world and that's how they would know the latest and greatest uh of capabilities. So we have our catalog here. Um and we have some retro features for some of our uh some of our customers um showing a little bit of the breadth of the nominal platform. Shinke

start with that one. So robotic, you know, fish uh processing the fish here live on the show.

A nice uh you know shining

uh Entures, you know, nuclear power.

This is amazing.

We've got Pratt Miller Motorsports, uh Hermus, Albido, Satellites, Planes, Regent. Um so yeah, um we've been sending these all over the the country over the last couple days. So

that's very cool.

Incredible.

Yeah. Uh how how have uh conversations with investors been around the SAS apocalypse? Everyone says uh I mean uh you know Toby look he was like I I vibe coded my own telemetry data. Uh is this not a threat to you? What's going on?

I think conversations um given that you're going to hit the gong again now. Um the conversations I think have been good. Yeah. I think um we actually see nominal I'll frame it this way which is um we're sitting at an incredibly interesting nexus of

like data coming off of hardware machines. Y

real human enriched engineering judgement like people are doing this work in our platform like they are deciding that that battery is good or bad and they're transitioning that sort of engineering judgment and insight um into the data and enriching it.

All of that is living in our platform and it is growing exponentially. So when you actually think about um if you believe this like world where models will longtail commoditize and it just becomes about interesting access to to data and sort of like modes around that um I feel really good about where nominal is positioned particularly there

and then I think we talk a lot about like what we've done in in our three and a half years so far I think has taken a set of an industry and a set of tools from 1990

to you know 2022.

Yeah.

Um and we've done that fast. Yeah. Um and now we have the opportunity and we've earned the right to basically bridge the industry into 2026 2027 right and so a lot of talking about you know what will we do with some of the funding I think a lot of it is like we you know we have our own sort of cutting edge we have our own AI team we're hiring and kind of building up and people focusing on how do we incorporate how do we both a nominal build with AI but also how do we um operationalize this for the industry um because right now going from you know a spreadsheet to uh to uh I don't know claw coding your own whatever is like um let's say it's not big

well yeah it's interesting I I thinking of it in a racing context the car is out on the track it's producing all this data you can imagine a world where an AI is able to more quickly totally like respond and help the team make decisions than any human right that's like kind of managing absolutely all these different

D I'm really I mean I'm really bullish on that I think like um we sometimes frame nominal a little bit as like we want to be and we think we will be like a part of one of um those systems of record. And so when you think about um when you're using we'll just use cloud code when you're using cloud code, right? Like it's still uh it's still interacting with like GitHub or whatever your version control system, right? Like is for for just sort of doing things like that. Um the same equivalent like doesn't really exist in this like software defined hardware world. Um and I think those need to be built frankly. Um

yeah I wonder I wonder how you think about like the economies of scale with regard to effectively software as a service because

I imagine that even if software the the instantiation of software the writing of software is commoditized uh there are still hard one lessons and almost secrets like you can't go to every LLM and say predict every problem that Yeah,

this drone manufacturer will have and then build software for that that will,

you know, immediately generalize to hypersonic flight or nuclear power generation.

It's a really good Yeah, it's really good. We're we're um we're working on some of this right now, I would say. And I think like a lot of our vision is the world of hardware testing uh right now in 2026 generally looks like you know setting out sort of a matrix of test points like very deterministic. My system my battery needs to be between this and this voltage under this condition

and then you sort of run that out. It it it happens very sequentially like that is how all of these systems are are tested and frankly that's how

talk about you know the the government that that's how testing happens. We are trying to bridge from a world of like what if you actually um can one do more and more in simulation and use outputs of pretty high fidelity often physics based models of the world. That's one input.

Um and then two what if you actually start to build like AI test agents

and their whole job is to determine the next best test point to run. So instead of running linearly through this matrix, like you're actually sort of in a kind of a threedimensional space where you uh are trying to kind of retrain and update the model and say, "Hey, if I could put this physical piece of hardware in any condition or state space, what would I put it in to optimize and sort of knowledge maximize my next step?"

Knowledge maxing. I like

Yeah, knowledge maxing. Yeah. So that's where I think that's where it's going. Yeah, I imagine like Shink likes that whatever you're learning from Anderol and whatever you're standardizing there applies to them. But then uh I was at Hermus uh the manufacturing plant and they had like a Prattton Whitney jet engine that I don't think I've seen any Anderl products that operate at that scale. And so whatever you're doing with Hermes is going to translate to Anderol and vice versa. And there's going to be there's going to be sort of like synergies there. Oh, totally. There's a ton of um there's a ton of compounding

where it doesn't necessarily make sense for every company to have their own system. No.

Um

I do you mentioned I think uh physics based simulation CFD. Um what what is off the shelf? What what bespoke systems uh continue to exist? Uh because it seems like the the messy Excel sheets, the messy Python version 27.3 maybe goes away. But um but what what actually sticks around uh long term?

Yeah, it's a really good question. In in the intermediate, we go to a lot of our our customers and they're sometimes shocked to believe that we can actually and this is not like necessarily a north star, but we can deprecate mat lab like just like

people can can get uh off.

And for the for the F1 fans out there, is mat lab a tool that you could use to like visualize the air flow over a race car? Or is it that

it's mat lab is more um it's sort mat lab is sort of like the og like you you'd use it in every mechanical engineering like master's degree it's just like it's it's all the yeah it's like just basically plotting graphing um but running like actual math um just engineering analysis like and it is um

so much uh of the post-processing of this like hardware generated data still happens in mat lab um mat lab's like a single player

tool by definition right and so I think one of the big value unlocks we've had is just like take a tool like that um and put it in the cloud.

Yep.

Um horizontally scale it so a ton of humans can access it and you can actually do engineering math and computation transformation in the cloud like that's um that's not I mean incredibly revolutionary itself but it's incredibly powerful in what it unlocks. So when you have to your point of like

um I think positive compounding effects between multiple different you know companies I think just within a place like Ander which is approaching you know 10,000 people the fact that um they're getting off of single player engineering and being able to do so much more in in multiplayer is a huge unlock. Then walk me through some of like the the Fed ramp ITAR compliance because

fun one

if I if I'm running a company and I want just a cloud hosted Jupyter notebook as multiplayer I could go to Collab like it's available but it's not designed with that use case in mind. It's a very general purpose tool, but walk me through some of the more uh defense tech government applications. And

yeah, so we we call it rugged deployability, but like if you if you come and you walk up and you say, "Hey, I want to use nominal." Um, we can deploy our entire stack fully airgapped on prem. Like that's something that we invested in from from day zero. And and obviously uh a lot of our customers have that use case.

What does that actually look like? Like

we'll ship you a server. Yeah, we'll just

Oh, you'll ship the hardware? Yeah, we have a bunch of servers in Austin. Yeah, we just central hub with

that. Is that like a Dell server with like a GP?

We have a deployability team.

Okay. So, you can kind of pick.

Yeah, we call them the deploy deploy boys.

Deploy boys. Okay.

No, so we um

because I I remember I talked to Sham Sankar at Palanteer and he said that the first time that they tried to ship a Palunteer binary to uh run on prem, the the the uh the person on the other side of the deal was like great, but this isn't running. and he was like, "You don't have nearly enough memory," blah blah blah. And that's the birth of the Ford deployed engineer project.

Yeah. Yeah. Yeah. I think it's um so we can we can with some customers it's easier. We'll just ship a tower. We'll ship a server to them. Um or we can just provide specs. I think I think people are the the department and and these like if the government is your customer, I think people are getting getting a lot better at that right now. So you can do that. We can also just deploy our software inside of our customers VPCs. That's often um and so inherit all of their you know security and credentiing and things like that. Um, but I do think that's been a big differentiator is like we've invested that would not normally be a place where I think you would spend valuable like startup engineering points if you were not in the industry that we are in. Um, but from day zero essentially we were like we needed to be able to deploy this. Um,

what about uh like approval cyber security compliance like I imagine that even if you have the best team of engineers possible like you still want third parties to work with that. What does that look like? Yeah, I think um you know there's a long long tale of of you know security and compliance um I think things there. Um I think you know we we're continuing to I think to work through that. Um some of the basics are like you know people have to trust that the tool is going to perform the the things that we we you know we say it is. And so in the early days we'd have lots of fun conversations with folks around like how are you doing like a low pass you know filter how like and just like explaining like basic math you know functions um etc. But we're well past that now. Um, so yeah, I mean I think we're we're continuing to burn down that. We we do like we support classified work, right? We have a facility clearance as a company. Um, a lot of our engineers are are uh are uh you know read in and I think all of the not all of a good portion of the

uh one like most interesting work with a lot of our federal customers is happening at at higher classifications and I think there's a direct correlation as you go to higher classification programs of like

uh a lack of tools.

Yeah.

Right. And so I think from a business perspective if we can get like there's there are really big budgets. Um people are you know they're I'll just give facts like I mean you know $3 million a day being spent on on test campaigns alone right and so if nominal is able to bring in some of those test timelines by you know a day uh and we can do more than that u it's like it's really valuable but when you go to some of those high-side programs um that are doing you know testing uh you really are restricted in what you can can use

crazy some of the most important work and the fewest tools.

Yeah. Is AWS sort of like the IBM of uh federally compliant cloud hosting these days? You remember IBM like no one ever gets fired for buying IBM. Is AWS still this leader in just like the default tool that you pull off the shelf if you want?

I think we see a lot of but we also see um it it really and this is one thing like I we see a bunch we see Azure we see other places like that. Yeah. So I think we we have sort of partnerships with all of the the the cloud providers um and and they all have their own efforts I think to provide you know some often some of the times like the the reason why we get slowed down is just because there isn't a environment for us to deploy the software in right like the customer wants it they they're asking for it um and we have to figure out ways to to get the software deployed but

yeah what's the team like where where you based

yeah so we are uh we're around 135 employees today we'll probably grow to around 240 um over the course of the calendar year. Uh headquartered in Los Angeles, uh Austin office as well, New York, DC,

and we opened an office in uh in London late last year.

London. Wow.

Yeah. So,

is that for international clients?

Yeah. Interesting. Um I went over to

I went over to London um probably six or eight months ago now. Set up a bunch of meetings um and kind

deal.

Yeah. And came back basically came back and all of those opportunities like materialized into real customer traction. I think we we sort of knew like this is not this is not a US-centric problem. This is a global like industrial sort of problem and there's a ton of opportunity and especially in more of the defense adjacent work unsurprisingly their uh the proximity that they're feeling is like is heightened.

Yeah. What's the mix of of business with uh companies that sell to the government versus government contracts directly? Are you feeling pull from the DOW directly?

Yeah. Yeah. So we we have um we have around 60 customers. Um

around twothirds of the work we do is is act is totally commercial. Um it's just a B2B engagement. I mean a company like Shinke or or Antaris etc. Um

and then we also directly work with the the federal government. Um and I always say

the biggest you know tester and validator of hardware in the world is is like the you know the department of war right now. Um, and so I think we're we're very happy to to support a lot of that that work and it's very needed.

Sounds great.

Gordy,

very uh insane progress.

Uh, where can people get these? Do do you just sell these? Do you ship them to people?

We've shipped a ton. We've shipped a ton out. I My sense is we will uh the reception's been very positive. Um, we'll probably be printing uh a lot more. So,

we're printing. I love it.

Yeah, we're printing.

Yeah, we're printing. Um, we'll be we'll probably be shipping them out. Um,

yeah. Uh what what is the meaning of Y cominator here?

Yeah, we we had fun with this one. Um so it's graduates who became astronauts.

Okay.

Uh Y cominator doesn't have it.

Oh no graduates.

We we support the US Air Force test pilot school

TPS. Some of the the coolest humans in the world doing amazing work.

We got to get Johnny Kim to go back through YC next thing.

We do. Yeah.

He's a legendary astron.

Yeah, we do. Hopefully that maybe that'll change one day. But yeah,

I would believe it. But if you're on the YC track, you're probably off of the

off of the astronaut track.

Yeah.

You got to build the next

to make decisions. Yeah. For now,

you got to build unserious software.

Yeah, for sure.

Well, thank you so much for taking the time to come talk to us. Always good to be here day.

I appreciate it. Congratulations.

Thank you guys.

Thank you.

And we will talk to you soon.

Okay.

Have a good rest of your day. Close out the show.

Yes. Uh thank you for watching today. Leave us five stars on Apple Podcast and Spotify. Sign up for our newsletter at tbp.com. Yeah, we didn't get enough into GBT 5.4

a.m. GT 5.4.

The reviews are incredible and we'll stack the beginning of the show tomorrow with updates there.

Lots of fun benchmarks to dig into, lots of graphs to analyze, lots of models to play with and examples to go and build or vibe code and see if they have that smell, if people like the the vibes. Anyway, thank you for watching. We will see you tomorrow. tomorrow. Have a wonderful evening. Goodbye.

Goodbye.