The Hidden Engine Behind CAD: Tech Soft 3D’s Big Bet on HOOPS AI

Episode Summary

Most engineers have never bought anything from Tech Soft 3D, and most of them use it every day. The company's toolkits sit beneath hundreds of CAD, CAM, CAE, additive manufacturing, construction, and PLM applications — roughly 700 companies in an industry small enough that everyone ships on the same foundations. In this special edition of AI Across the Product Lifecycle, Michael Finocchiaro talks with Jonathan Girroir of Tech Soft 3D about the company's move into AI with HOOPS AI, a framework built to make rich 3D engineering data usable inside machine-learning workflows.

The conversation is a useful corrective to generative-design hype. Girroir argues the most valuable AI in engineering is not a flashy demo but the ability to find, understand, and reuse the knowledge already buried in decades of product data — geometry search, feature recognition, and part reuse. Getting there depends on unglamorous plumbing: engineering workflows moving from files to APIs, cloud adoption settling into hybrid architectures with real data-sovereignty constraints, and native multi-CAD data mattering more than neutral geometry, because triangles alone strip the context that makes data trainable. He is openly skeptical of text-to-CAD, and candid about build-versus-buy for startups: build your UI and your web portal, do not build a solid modeling kernel.

For PLM and engineering software teams, the strategic thread is where the glue between systems still is not — the digital thread remains, in Girroir's view, the most underestimated opportunity in the industry. The episode also covers whether OpenUSD could become a common layer across engineering and manufacturing, the coming Tech Soft 3D Data Hub, whether engineering gets its own “OpenAI moment” before 2030, and how quantum computing could reshape simulation. Girroir's honest read on the timing: this industry moves slowly, and it always has.

Tech Soft 3D quietly powers much of the engineering software industry.

Its technology sits beneath hundreds of CAD, CAM, CAE, additive manufacturing, construction and PLM applications. Now the company is making a major move into AI with HOOPS AI, a framework designed to make rich 3D engineering data usable inside machine-learning workflows.

In this special edition of AI Across the Product Lifecycle, I speak with Jonathan Girroir of Tech Soft 3D about:

• Why engineering workflows are moving from files to APIs • How cloud adoption is producing hybrid architectures • Why native multi-CAD data matters more than neutral geometry • Where HOOPS AI fits into geometry search, feature recognition and part reuse • How CAD data can connect with costing, procurement and manufacturing information • Why text-to-CAD is still overhyped • Whether OpenUSD could become a common layer across engineering and manufacturing • The coming Tech Soft 3D Data Hub • Build-versus-buy decisions for engineering software startups • Whether engineering will experience its own “OpenAI moment” before 2030 • How quantum computing could reshape simulation

Jonathan also explains why the most valuable applications of AI may not be flashy generative design demos, but the ability to find, understand and reuse engineering knowledge already buried inside decades of product data.

Timeline

00:00 Introduction 00:37 What Tech Soft 3D does today 01:21 HOOPS, SpinFire and 30 years of engineering software 02:41 Cloud, hybrid architectures and data sovereignty 05:30 Connected workflows and engineering ecosystems 06:33 Why rich multi-CAD data matters 07:51 What customers want from Tech Soft 3D 09:29 Rendering, OpenUSD and interoperability 10:53 Major 2026 announcements 11:09 Introducing HOOPS AI 12:24 Is text-to-CAD overhyped? 13:31 Geometry search, reuse and contextualized 3D data 15:01 SpinFire mobile, cellular volumes and the Data Hub 16:32 The customer problems behind the roadmap 18:31 Why HOOPS AI generated the biggest reaction 20:15 Supporting engineering products for decades 22:43 What customers really want from AI 23:26 Where AI creates genuine value 24:04 Where engineering AI is overhyped 24:46 Why topology, PMI and design intent matter 25:49 Visualization across the enterprise 27:59 Web-based 3D collaboration 28:40 Build versus buy 31:02 Openness, interoperability and IP 33:36 Advice for engineering software startups 35:17 How Tech Soft 3D supports startups 35:49 Will engineering have an OpenAI moment? 38:45 Quantum computing and simulation 40:15 The opportunity in the glue between systems 40:37 What comes next for HOOPS AI 41:16 How to evaluate Tech Soft 3D 41:42 Closing thoughts

#CAD #EngineeringAI #EngineeringSoftware

---

Full Transcript

Michael Finocchiaro

Welcome to the AI across the product life cycle, a special edition with my friend Jonathan Cher of Techoff 3D. How you doing today, Jonathan?

Jonathan Girroir

Hey, Michael. I'm doing great. Thanks. It's great to be here.

Michael Finocchiaro

Yeah, we've actually crossed each other numerous times this year already in trade shows and so forth. It's been great meeting you in person. Um, Texos 3D has been quietly powering engineering software for decades. For those that only know the logo, what is the company about today?

Jonathan Girroir

Yeah, we are primarily a software development toolkit provider for the 3D engineering industry. So we have SDKs that provide data translation and visualization, simulation and analysis that power probably almost all of the the independent software vendors out there in one way or another. We we touch so many in our industry across the whole product life cycle. So from um design to simulation to CAD, CAM, 3D printing and then also in the building construction space. And we also have a small part of our company where we actually use those toolkits and we build end-user products for uh visualization in the enterprise.

Michael Finocchiaro

Well, you were you're talking about the hoops the pretty much the industry standard for visualization, right?

Jonathan Girroir

Yeah. Yeah. People know we're the Hoops guys versus uh being techsoft 3D is Hoops is even more recognizable brand. So that's our our set of SDKs, but we also have the Spinfire brand. That's our engineering um visualization tools for the enterprise, data translation and and uh AR and VR uh end user tools as well.

Michael Finocchiaro

Awesome. Um yeah, and it also has been a bit of a transition for you guys, right? Because you had a very one of the very famous people in the CAD world for a long time, Mr. Fritz, uh who uh stepped away and uh but you guys have done a great job picking it up and and and keeping the ball rolling. In fact, you told me business is booming. So that's awesome.

Jonathan Girroir

Yeah. You know, when you talk about the history of Techoft 3D, we've been around for almost 30 years. We celebrate 30 years in November of this this year. And in fact, we we spun out of Autodesk. So, so Ron is is known as in the industry. Um, originally Carl Bass, who uh became an executive at at Autodesk, was also an original Hoops Hoops guy as well.

Michael Finocchiaro

Yeah. He's famous for saying PLM was useless, right?

Jonathan Girroir

Yeah. Lived. everything he said was correct but but is a recognizable name.

Michael Finocchiaro

Yeah. Um so what trends are you seeing from from your uh from your seat as a visualization guy in terms of trends across CAD CA PLM manufacturing that are reshaping expectations of the customers?

Jonathan Girroir

Yeah, a number of different things. And so um one move to the cloud a lot of people are like what like that happened a decade ago but in our industry we kind of move slowly especially some of the bigger players. And so that's still very topical is how do we move legacy applications to to the cloud and become cloud enabled. Even simulation and analysis becoming um cloud enabled uh CAM being cloud enabled uh CMM being cloud enabled things that really were resistant to that initially. So we're seeing we're seeing that um and then the the build of of hybrid architectures too. So cloud and desktop. So what workflows belong in the cloud? what still need to be on the desktop and kind of what's the best choice for different workflows. So that's one. Um we see a lot of data. So so really at our core here at TechSoft, we're we're a data access and translation um company. We have visualization on top of that. We have simulation uh with that as well. But it really starts with data and the way in which people are interacting with data, sharing data and moving data through workflows um is becoming less file-based and more API based. So as things move to the cloud, access to that data becomes not just file-based but really just holistically looking at how do I access this through a set of APIs. So we we really are seeing a a evolution a revolution in in that space. And then of course AI, right? We're going to talk about that quite a bit today. Um, but with all of that data, how do we how do we use that to make intelligent workflows and and learn from um connecting that to business information um and and extract value from that using uh some of the newer machine learning tools.

Michael Finocchiaro

It's ironic you mentioned that because it seems that um in my talks with uh some of the startups and customers it there also is even a wave back to onrem again as of the concerns on sovereignity and so forth. So we all had this big pendulum towards the cloud and it's sort of starting to swing back a little bit which is a bit interesting right?

Jonathan Girroir

Yeah. And everything in between. So CO definitely opened up a lot of workflows just because we had to be remote and so that kind of was the last that's what we needed to kind of move stuff um to the cloud.

Jonathan Girroir

Onrem is still absolutely needed to be supported and and private cloud and government cloud there's there's a lot of different solutions on how to own your your data and still maintain it yet create this connected environment.

Michael Finocchiaro

Um oh speaking of workflows so it feels like engineering software is becoming less about indiv individual applications more about connected workflows. Would you agree with that statement?

Jonathan Girroir

Yeah. Yeah. Absolutely. And it's it's about moving data from application to application. It's about it's about like ecosystems and making sure you play within that ecosystem and and big vendors are trying to lock you in and build massive ecosystems where they have all of the tools that you need. Um but then everyone else is trying to figure out how do I tap into that? How do I make it as seamless as possible? And so yeah, over the years just standard file um formats that become richer that allow you to have these integrations and these translations like that's that's definitely something that we're seeing. Um, but then direct integrations through APIs that that is also really really important because you don't want to lose your data, you want to lose your work and there's a lot of value and and connected workflows for the whole product life cycle. Yeah, absolutely.

Michael Finocchiaro

And uh what capabilities do ISVS need today that were not even priorities say five years ago other than AI?

Jonathan Girroir

Okay. Right.

Jonathan Girroir

Yeah. Um multicad enablement is really what it comes down to. And so like not being able just to exchange a step file, but be able to read like let's let's get the Creo file or the NX file with all the PMI and the feature data and the feature tree and and all the metadata and material information and kinematics. So just that that real rich data access is is important and then be able to use it, right? So not just be able to bring the file in but then make informed decisions on top of that. Um probably go ahead

Michael Finocchiaro

probably like unique ids too so that like if you're bringing your data through a series of steps you can kind of track changes

Jonathan Girroir

and as you implement something and change it you're able to implement that change as well like if you're doing a draft um draft angle analysis but then you change some of your geometry all of your preparatory work can get reapplied so so those are really important as well

Michael Finocchiaro

and I suppose especially in your case it's also work horizontally in the organization but vertically up and down the supply chain, right?

Jonathan Girroir

Yeah. Oh, abs. Absolutely. Yeah.

Michael Finocchiaro

So, so what are they customers asking for the biggest improvements? What are what are the what's the customer demand like? What are they asking Techoft 3D?

Jonathan Girroir

You know, more data. I really is is there's still a lot of vendors out there trying to keep their own little corner of the ecosystem. So I know in in recent months we've added support for um DGN file formats making sure that you can just natively bring that data in be able to visualize it and work with it seamlessly with without having to convert it to a neutral format. um speed I mean that it's data our data is getting bigger every day the way we acquire it the way we model it the way we visualize it and and computers are getting better of course but speed is always important so that's the speed of load speed of translation speed of visualization speed of interaction speed of processing so so being able to continue to accelerate that um be able to parallelize the processing so now if everything's on the cloud how do I actually make use of these parallel environments and and renewed parallel environments or on or on your desktop as well or being able to accelerate on the graphics card. So, so I guess like rich data is something we continue to focus on and then we continue to focus on speed and and to some extent v visual fidelity. So, um, photorealistic rendering, ray tracing is still something we see with with some of the newer tools like nano banana like we're able to push data there and get pretty pretty good high high-end rendering from it.

Jonathan Girroir

But, um,

Jonathan Girroir

PBR, it's it's um, physically based rendering is a slightly new standard for fast graphics for like high fidelity. So being able to support that. Um and then like there's some new paradigms of file interoperability that we're looking at that are quite promising like USD universal screen description.

Michael Finocchiaro

I was about to ask about open USD.

Jonathan Girroir

Yeah. Like that's really exciting because you that came from Pixar but now it's being adopted by Autodesk and Apple and Adobe and Trimble and there's just a ton of momentum and what makes it so humans.

Michael Finocchiaro

Yeah. and and you saw this uh the twin thread acquisition by u ava was particularly interesting about a month ago because that's what Andy Wott had bet his entire company on was open USD.

Jonathan Girroir

Yeah, a lot of people are doing that and it seems to be paying off and we are as well. So we're heavily heavily vested in that and and a a member of OpenUSD and trying to support that format and inform it. What makes it really exciting is um it's like streamable, it's composable, it's extendable, it's very flexible. Um and and because there's industry momentum around it, um we're able to extend it. So they're like they're creating their own brep representation, but you can you can plug parasol data in there if you want to because it's it's very extensible.

Michael Finocchiaro

So let's move on to the recent announcements. You can tell us all the the wonderful stuff you guys been doing. And I know we've you've come to my trade shows and been able to talk about it. So can you talk about some of the more important uh initiatives that you've done recently? Walk through what was announced. Why did you announce it now?

Jonathan Girroir

Yeah, I mean the big thing this year 2026, we released officially Hoops AI, which is a a framework for taking CAD data, 3D data, and being able to encode it um and be able to take it into machine learning algorithms. Um, so it's a framework for doing that. We have uh some encoders that allow you to quickly get it into different frameworks for machine learning. We're also able to connect it to business information. So, some of our our more recent releases allow you to connect it to costing information and um create informed businesses decisions around that. So, we're we're all in to AI and um particularly what's what's really important about this is it's based on the 3D geometry. So, it's not like a a 2D representation that we're we're feeding into an algorithm, but we're actually looking at your topology and your your 3D geometry. So, we're the the only um company that's that's currently doing that and providing this as a toolkit um that then uh can be built on top of. So, trying to help all of our customers who are using our translation tools to be able to now take all of their data, encode it, and then start to use it in these these um new paradigms.

Michael Finocchiaro

Well, this one's sort of off script, but I wonder what what's your opinion on the uh um the these text to CAD uh startups that are popping up like mushrooms, right? I think I read about a new one every other week.

Jonathan Girroir

Let's go. Let's go. Um I, you know, that's that's the holy grail and and a lot of people are trying to get to that. Um we'll see. And I think it what what it comes down to is really good good data, a good, you know, what what's made a lot of these predictive um tools outside of our industry uh work is because they have a large body of work that allows them to make these predictions. And so building on top of that, you you need good you need good data and you need a fair amount of it. And so I'm curious to see the few that I've used are just kind of like uh like let's let's see and it's going to it's going to take some time. Um what we're really interested in is accelerating existing workflows and making making large libraries um accessible large libraries of 3D data accessible and navigable. So for instance, if I have this widget that I might have I'm an aerospace manufacturer. I might have designed it five times already. I have in my library, but how do I find it and how do I connect it to business information? So it already has um let's say the FAA approval and it's gone through a number of steps and it has this analysis data and we knew we know how to how to build it, how to manufacture it, um all the testing information. And so if I'm able to take this part and find it and not have to re-engineer it um or find similar parts and be able to use those, that's that's incredibly valuable right now. And that's that's where we're trying to work in is is part search and and feature recognition. Um text to CAD, we'd love to support that as well through our translation tools being able to provide that to you. Um, but right now we're just trying to bubble off to smaller chunks of of foundational uh approaches.

Michael Finocchiaro

It's interesting the way you phrase that because it sounds similar to what a lot of the innovation I'm seeing on the manufacturing side of the fence where they're taking the IoT, the scattera, the PLC's and recontextualizing. So that's what you're saying. You're taking the 3D and recontextualizing it in the PLM engineering context basically, right? Whereas open USD do that but also contextualize in the manufacturing context which is what makes it unique.

Michael Finocchiaro

But I don't think those are the only announcements you've made the AI ones. Weren't there some announcements around spin uh uh spin

Jonathan Girroir

spin fire. Yeah.

Jonathan Girroir

So yeah we we released a mobile viewer recently. Um we like I mentioned we supported DGN. um cellular volumes for simulation is huge and yeah we talk about CAD and CAM but simulation analysis is a bigger industry uh dollar for dollar and so making sure that we're able to visualize cellular volume so it's it's uh a way of which you can be looking at simulation results or or pre-processing data so making sure that in our graphics engines we're supporting that um we're starting to talk about a data hub. So that's really exciting. So that's kind of future looking later this year. So um kind of a think of think of allowing your opening up your data throughout the engineering enterprise through a set of APIs and and through a set of of plugins and tools that sit on top of that. So so there are like a number of things happening happening here at at Techsoft. Um USD we've been announcing the support for that. Um and then and then also later this year there's another format for um quality assurance and and coordinate measuring that's QIF quality I forget what it stands for but

Michael Finocchiaro

interface format or something.

Jonathan Girroir

Yeah. And that that's kind of like a very structured powerful way of of bringing all of your MBD data in and and then also including inspection results from a CMM machine

Michael Finocchiaro

or how to do inspection as well. So, so what kinds of like specific customer problems uh were you trying to address by some of these announcements? So, there were particular things where people said, you know, I really have this problem and darn it, if if Techoft 3D could give me something.

Jonathan Girroir

Yeah. I mean, we we went in a bunch of directions there. And so when we think about how do we invest at TechSoft, so the main thing is making sure that we're current um and unlocking data throughout the enterprise for our our customers. So we have about 700 different independent software vendors that are using our tools and they need to be able to access KIA, Solid Works, Creo, NX data, you know, 30 different file formats um within their application. So we're we're multicad enabling them and a fair amount of our energy just goes into making sure that we we maintain those translators. So that's that's a big thing. On top of that um we do look at you know where are our customers going when it comes to uh features and and how how do we support them in in doing that. So new formats like USD and QIF and DGN. Uh and then on top of that like where is innovation happening at Texop? So that's that's coupoops AAI and and that you know that's you know there's there's little things like uh I forget like configurations in in in Creo for instance or multi-threading uh KIA file translation. So there's always these these little things that we're trying to improve but right now a lot of our R&D is just going into to Hoops AI um as well as our data hub. So, I guess those are the two major initiatives here at TechSoft is is making sure that we're we're ready for that final migration to the cloud. The opening up of data through APIs for MCPs um is is really interesting to us and and then also um yeah, just Koops AI and and being able to encode your CAT data.

Michael Finocchiaro

So, what um which of the announcements you mentioned had the mo biggest reaction from your partners?

Jonathan Girroir

Hoops AI.

Michael Finocchiaro

Yeah.

Jonathan Girroir

I mean just when we when we look like this is a small industry. How many independent software vendors are building 3D engineering um applications? I mean you you know that better than

Michael Finocchiaro

123 about just design intelligence another 200

Jonathan Girroir

but it's a small it's a small industry right and so we have about 700 different companies using one or more pieces of our technology. Um, and so when when when I host a webinar, like I'm excited to get a hundred people there. Like that's a good day for me because you know how many how many software developers are doing that. But when we released Hoops AI, we saw we had to we had to upgrade our our Zoom webinar account.

Michael Finocchiaro

Wow, that's awesome.

Jonathan Girroir

Um, and and so the response was great. We lead with that. Um, and it opens doors. We're finding a lot of a lot of companies are like they like the idea of AI. They're trying to figure out, everyone's trying to figure out how do we how do we make appease our board, our investors, or you know, how do we stay competitive? They don't quite know how. They don't might not have the internal experience with machine learning. Um, but it it intrigues them. And what we actually find is they might not be ready to implement machine learning algorithms, but they still need data translation or they still need graphics. And so even though it's interesting to them, we actually uh find that we start we start at the very base level is like all right, let's get you some some high quality data access and then we can start talking about about other um more complicated workflows and other toolkits on top of that

Michael Finocchiaro

and and some of these new capabilities are they completely complementaryary to the current offering or some of them are or are you also suns setting things along the way?

Jonathan Girroir

Uh we have an ecosystem So I think we like to talk about a we have about 10 different SDKs that that we um make ourselves and then we also partner with companies like Seammens and Machineworks for their modeling kernels. So Seammens has Parasolid, Machineworks has their Polygonica mesh mesh modeling toolkit. Um so that ecosystem everything kind of works with bridges and and connected APIs um embeddings with within them. Um, do we sense do we sunset things occasionally? Uh, trying to think what's what's been sunset in the past. Um, maybe like a we had a ray tracer that we're not we're not actively pushing right now. Um, but we're we haven't even sunset that. I'm trying to No, I we've consolidated. So So we we're trying to simplify our product offering actually. So we we had uh we had a 3D PDF creation and kind of a file writing set of tools that's just been incorporated into Hoops Exchange. Now that's our file access. So it's read write. So you can read CAD data in and publish a 3D PDF. So we what tends to happen is we are in we we build relationships with companies and they ship their products on top of our our SDKs and the life of those products are measured in decades and so it's very hard for us to sunset technology

Jonathan Girroir

you know may maybe like the the biggest sunset news is like we we're not supporting Visual Studio 2015 for instance like I mean that's really what it comes down to or or if if um If Apple's kind of if they sunset let's openGL for instance um then then we try to respond to that. So Micros so the big vendors when they sunset a technology we have to pivot and those those are the really the the big things that and same time like we still have to support

Jonathan Girroir

even if they sunset we still have to support. So, so we try to be this real stable foundation with a long-term partnership and and literally we have we have customers that work with us for decades, which is great to see.

Michael Finocchiaro

That's awesome. Um, but let's step into the like the heart of the the debate which we we've already been touching around about AI. So, everybody's obviously talking about it, but what are your customers what are Texoft customers really asking for today? What is their their main ask for when they come to you and say we need AI?

Jonathan Girroir

Yeah. Uh there's they're still trying to define that. Like everyone seems to be building an agent for automating the UI within their application. Um but where we're finding some real value is connecting business information. So that could be procurement, costing, manufacturing, um quoting, connecting that to the original CAD data and um focusing on geometry search and feature recognition are kind of like the big areas that we continue to see themes in this space.

Michael Finocchiaro

And where do you think that AI is creating genuine new value inside traditional engineering workflows? Where you think is the real the wedge that it's creating in there?

Jonathan Girroir

Where is it creating value? um automating wrote tasks. I mean that's that's absolutely true. Uh and then pro providing insight into connecting the dots. Um so for instance like just not not knowing what you have available to you and and being able to organize it in a meaningful way. So these these parts catalogs being able to bring some some sense to to them and organization to them as well.

Michael Finocchiaro

And where do you think that we're overhyping it?

Jonathan Girroir

Everywhere.

Michael Finocchiaro

Well, specifically in in engineering and industry, what do you what things do you think we've been talking about and you're just like, "Yeah, right. Right, whatever."

Jonathan Girroir

I I mean, I'm still text to CAD. I'm I'm think with anything, we'll get there. But I think like that's a long ways off, honestly. Or like Yeah. Like I these

Michael Finocchiaro

I thought I thought you were going to say pens, right? Everybody's

Michael Finocchiaro

bragging on pens nowadays. You know, the physically informed neural networks are sort of out of out of out of fashion now.

Michael Finocchiaro

Yeah. Um, but what what does how does and you've already touched on this, but I wanted you to be a to re re uh you know nail put the nail on the board on how does high quality engineering data particularly 3D data become a real advantage for for AI.

Jonathan Girroir

Yeah, the more data the better. And it's more than just triangles. It's more than just bre. It's more than just topology. But when you're able to look at the PMI and the relationships within assemblies and the metadata. So, so this is all encoded into our systems these days in our in our data and our files. And so making sure that you have access to that um is is is really really important

Michael Finocchiaro

contextualization of of the thing.

Jonathan Girroir

Yeah. And then then back to the workflow, it's like being able to to actually track the data back to, you know, how is how is it and why was it built that way? Of course. Uh well, design intent, right?

Michael Finocchiaro

Yes. Yeah. Um yeah, that's a really huge field at the moment in terms of startups uh coming up with uh interpreting and and um interpreting design intent is is absolutely huge. Um in terms of visualization collaboration which is really one of the core values that you guys provide um engineering teams today they include suppliers manufacturers service uh procurement you know even end customers how how is that changed the requirements for visualization particular probably it's forced you to have really simple user interfaces because you're going to have a lot of people that aren't technical right

Jonathan Girroir

yeah and so we we talk about um engineering data throughout the enterprise so not just on the designers, the engineers desk, but throughout the enterprise. And we're trying to unlock that in a number of different ways. So, so we do that two ways. So, our SpinFire set of products, those are end user products, um, allow you to have lightweight viewers on mobile, desktop, web, um, throughout your enterprise at a kind of a lower cost, but having rich data access. Um, so we we do that. That's super attractive to to people. And then at the same time, which is interesting, we compete with um our customers on the toolkit side of things. They're they're doing that in in unique ways as well. And that's that's really important. Um so having having it lightweight, having it streamable, having it on a mobile device or or in a in a web portal is is really really important. People want that um because it helps make informed decisions. And so

Michael Finocchiaro

I was actually about to ask that question on how important are lightweight data formats. So

Jonathan Girroir

very

Jonathan Girroir

yeah I mean the flavor of the format changes right so you can you can stream stream geometry there there are lightweight formats out there there's lightweight representations that's that's why USD is interesting kind of bringing it back to that is is it can sit on a server and you can just request let's say the visualization data without all of the the heavy metadata and you can request that um on demand and so so yeah We do see a lot of opportunity uh there as well.

Michael Finocchiaro

So what roles do these web technologies for 3D now play compared to the way we used to just use desktop tools?

Jonathan Girroir

Collaboration review. Yeah. Yeah. Absolutely. These connected environments. So teams, distributed teams, right? So that so we're we're global workforces. And so making sure that you just don't have it on the the shop floor or or in in the floor of engineering, but you have to be able to to have a a connected system that spans continents as well and be able to collaborate in real time or be able to check in and provide feedback and um that's that's super important as well.

Michael Finocchiaro

So software vendors are increasing, excuse me, I'm out of water. Oh well. Uh software vendors are increasingly wanting a best of breed components rather than building everything themselves. Are you seeing that um accelerate that the fact that they're they're actually building with uh you know with external tools or do you think that we still have this cons you know uh one-stop shop kind of a attitude?

Jonathan Girroir

Yeah, I the build versus buy I think is is the the the term there. Um we actually have a ebook coming out pretty soon. Maybe we can share that.

Jonathan Girroir

Yeah, we're we've done some research with tech clarity around that. There's a lot of reasons to drive.

Jonathan Girroir

Sorry. Uh

Jonathan Girroir

but there's a lot of reasons to build versus buy.

Jonathan Girroir

Um you should not build

Jonathan Girroir

data translation. I'm just letting you know that you should not build a solid modeling kernel. Um,

Jonathan Girroir

but you should you should build a a UI, you should build a web portal. And in some ways, like graphics are kind of a commodity now. And we're kind of seeing that sometimes they you people will use our graphics, sometimes they'll build their their own graphics. The benefit of using components is they're integrated. And so like our graphics are very very focused on engineering data and navigating it. So, there's some real benefits there. Um, but then buying an off-the-shelf product like our SpinFire viewers, there's it's quickly you can deploy it quickly. They're secure. You don't have to maintain them, but when you buy something, they're not configurable. They might not fit your workflow completely. You might have a very specific workflow and intellectual property to the way in which you do design manu and manufacturing um which would require you to build. So, so it's like this either or. Um, but some of those found foundational components like data translation, solid modeling, graphics to some extent, simulation as well, simulation engines and accelerating simulation. Like a lot of that stuff um you'd want to get a component for or just buy off the shelf.

Michael Finocchiaro

I'm not going to go build anis.

Jonathan Girroir

No. Yeah.

Michael Finocchiaro

Well, actually there are people that trying to

Jonathan Girroir

Yeah, I guess so. I guess there's always there's always a exception to the case, isn't there? Yeah.

Michael Finocchiaro

So, how do you balance openness with protecting intellectual property? Which is an interesting question, right? Because that's, you know, the I think one of the pivotal moments in the CAD industry was when Chuck uh Grin and Tony Auso said if we sell our kernel to everybody else, it's just going to make us work harder to make better software because the differentiating factor will not be the kernel, it'll be the software around. So, how do you guys balance that? because you guys are also have an open toolkit and yet you want to make sure that the the best stuff is uh you know people are coming back to tech for it.

Jonathan Girroir

Yeah, you know um we we're very open because we we open up file formats to everybody. So that's like the first thing is we're in an openness enabler where you can read any CAD data for the most part and be able to access it within your application. So that's like that's a big deal. Um we don't open source it. That's our intellectual property. So you know those binaries are are specific to techsoft but we are trying to facilitate openness because of that and provide it the same thing is we will provide it to everybody. So we provide it same technology to seammens as we do to so as we do to autodesk as we do to the startup to to endop um and so that's that's really important. We're not affiliated with any of those companies which which could you know if you buy a component from them um they may revoke your privilege if you become too competitive. We are the Switzerland of of CAD uh component technology which is a great place to be um and have everybody benefit from that. There are some things like we open source our our um our UI for web webc viewing our our viewer there. So we we find that that's that's um really important and we're we're always asking ourselves like this is a big question is like should we just start giving away some CAD translation for free um and just so so that's like it's an active topic of of tech soft like trying to figure out you know where can we continue to support the industry not all CAD translation but kind of what areas and supporting open standards so open standards like IFC for building and construction open standard there step

Michael Finocchiaro

step for yeah for all of its kind of um

Michael Finocchiaro

you know for the the quality stuff

Jonathan Girroir

yeah US USD JT which is like pseudo open so we we support JT and and partner with with Seammens who also kind of has this level playing field approach to to components as well

Michael Finocchiaro

thank you sem components our friends Evan and Evan and company um so what advice would you give to startups when they're in that um build versus buy kind of thing? Um I mean your your answer is obviously going to be well you know when it comes to visualization you come to us and you don't bother. Uh

Jonathan Girroir

it it depends. Um like I said visualization you you may be able to distinguish yourself there. Like I know we have like point cloud visualization for instance you might be like really good at that. um just do what you're good at and then leave everything else foundational up to a component provider. So particularly data translation, solid modeling, mesh modeling, um visualization to some extent honestly what I would recommend if you're a startup, you you have to create an MVP as fast as possible, right? You need to show a vision and so go use don't go build a graphics engine. go use one off the shelf and if you want to distinguish yourself in a unique way later that's fine but like you need to create a product it's and today with um the way in which you can use a co-pilot or or other AI coding tools like it's not you don't pitch a PowerPoint deck anymore you pitch an application and so you build you build something um and you want to show it and these SDKs allow you to do that incredibly quickly and most of them have a certain evaluation phase or a startup program that allows you to use the technology and get funding and then we can talk about how do we how do we enter into a commercial agreement to use that technology?

Michael Finocchiaro

Well, let me let me just tack on to that question. So, how um how does it work in terms of your your particular work with startups? How do you guys enable startups to add this uh the visualization stuff of techsoft into their software without it becoming an additional burn on the limited amount of capital they have?

Jonathan Girroir

Yeah. Uh we give a free evaluation and then we also couple you with uh content expert an engineer so engineer to engineer support to implement that as well and make you as successful as possible.

Michael Finocchiaro

Very cool because I know you do have quite a few startups on the list that are uh customers which is great.

Jonathan Girroir

Yeah.

Michael Finocchiaro

Um so looking ahead I know it's kind of ridiculous the these questions are always three to five years and we can't even look ahead six months anymore but um well actually I'll change the original question. Do do you think we're going to see an open AI moment? I just had a call with two French startups about an hour and a half ago, uh, Oplid and Cognix, and they were like really bullish. They think we're going to have an engine open moment, certainly before 2030, maybe even before 2028. Um, what what do you think about that? Do you think u how close we are to before and after in engineering? Because we certainly aren't there yet, right? We're still in the old world for the moment.

Jonathan Girroir

Yeah, I our industry just moves slowly. It just it just does. I I would love to see that, but even some of these what you call startup unicorns, it's it's been a five to sevenyear just grind for them.

Jonathan Girroir

Um

Michael Finocchiaro

like the ants or the uh or collab or guys like that.

Jonathan Girroir

Yeah, absolutely. So, so our industry moves slowly. Um I don't I don't see a massive paradigm shift. So like for instance um uh convergent modeling which is tessellation plus topology or hybrid modeling that was a big thing 10 years ago and like we're just starting to see workflows adopt that like it's it's a good technology and then um with endtop what's what's their their modeling kernel their

Michael Finocchiaro

implicit implicit

Jonathan Girroir

implicit I mean there's very unique good uses for all of these but is it going to flip our industry upside down I just don't I just don't think so.

Michael Finocchiaro

Um, I I wish I was going to be in Washington DC next week to see uh CDF fam because

Michael Finocchiaro

I don't know if you read about it and th the other people listening hope somebody hopefully will go to Washington next week and then tell us about it but um they're going to do a demo with six startups and one demo endtop is digital ciscuit C infiniti and I'm trying to remember the other two and it's going to be apparently it's going to be pretty close to an open air moment in terms of infinite design spaces. and you know being able to do a requirements based design uh straight through including all the security with air gaps. It's going to be I think it's be kind of crazy. I just it's a long flight to America just for me from Paris anyway.

Jonathan Girroir

Well, let's go. Prove me wrong.

Jonathan Girroir

Yeah. But um I I do feel like things are happening faster than ever and and the the rate of innovation and and and our industry is feeling that and that's going to continue it's going to continue to accelerate. So that's absolutely true.

Michael Finocchiaro

Okay. Well, so so if we if you were to imagine looking 3 to 5 years ahead, what changes you think would be the most surprising looking back to software engineering?

Jonathan Girroir

Yeah.

Michael Finocchiaro

Engineering. Software engineering. Sorry.

Jonathan Girroir

Engineering software. Um I mean one thing that is what's the next paradigm shift looking towards that. So you know AI we're riding that wave. We're still kind of I feel in its infancy there. So that's that's going to run its course. Absolutely. Um what I am interested in the in the future maybe even five more than five years um is is quantum computing and how that's going to impact our our industry particular question. you know there's certain workflows that can be accelerated particularly around simulation right so that's a data heavy space and and easily paralyzable and so we barely have the compute p we don't have the compute power yet at all but you know that's going to be changing multiple industries and I think I think ours as well and so kind of what does that look like and and and how do we support that and what workflows can we migrate there

Michael Finocchiaro

I do have several startups including BQP and consent that are looking very Yeah, Quadrant just closed a massive uh funding round uh not even two months ago. Um I I'll probably I'll be seeing them in Elzinki. I'll be going to the Slush conference in November. So I'll be able to see some of my Finnish friends up there. Um if you were starting an engine company today, where would you focus?

Jonathan Girroir

I think that's where I would where I would focus.

Michael Finocchiaro

Quantum still there.

Jonathan Girroir

Yeah.

Michael Finocchiaro

And which technologies do you think are still underestimated today? What are we what opportunities are we missing? what rocks should be looking under

Jonathan Girroir

the glue between systems,

Michael Finocchiaro

the digital thread.

Jonathan Girroir

Yeah. Yeah. That's still there's still opportunity there.

Michael Finocchiaro

Um so we're almost at the end of the questions. Um I wanted to ask uh what's next for over the next 12 to 18 months. What can we expect from you guys?

Jonathan Girroir

Yeah, a doubling down of of um different tools that sit on top of Hoops AI. So um being able to connect that to business intelligence is is going to be real important. And then the release of our data hub uh technology that allows you to deploy quickly enterprise scale through APIs not not file based but API based data access.

Michael Finocchiaro

And if nobody ever someone that's uh listening has never evaluated your your products before where should they start?

Jonathan Girroir

techsoft 3d.com and navigate to any of our like evaluate now pages and and we can set you up. Um download the product, get the learning material, get connected with our team. Like I said, we have a a global team that's able to support you engineer to engineer and and that's something we really enjoy is being able to understand your problems and and help you build solutions for those.

Michael Finocchiaro

Awesome. Well, I've had a a great time today. I appreciate the uh

Jonathan Girroir

super fun

Michael Finocchiaro

all the answers. Do do you have any uh closing comments before we go to the audience?

Jonathan Girroir

No, I things are happen faster than ever. That's really absolutely the case. Um and you know, it sounds like you got quite a big travel schedule. I I am really enjoying connecting with people either either virtually or or in person. And I feel like in this very disconnected AI world, um the it's the personal connection that I'm I'm really enjoying. a lot more. Um, so looking forward to seeing you and and other and your listeners uh wherever we may be later later this year.

Michael Finocchiaro

Are you going to be at ITMS in Chicago?

Jonathan Girroir

Absolutely.

Michael Finocchiaro

I haven't I haven't decided on that one yet, but I do have slush on the agenda. Um I want to do another threaded. I haven't figured out when or where yet. That's that's a work in progress. Um but we'll find well certain maybe maybe Chicago then. That'll be fun. Have a a beef sandwich and some beers. Yeah.

Jonathan Girroir

Pizza. deep dish pizza. Yeah, sounds good.

Michael Finocchiaro

Before we go, actually, my friend Brian Carol, who's a regular guest on my other podcast, is asking, "What is your opinion um about the profile that's forming referred to as product memory?" You probably see some of the debates on that with stuff.

Jonathan Girroir

I'm not too familiar with that, so I I can't can't comment on that, unfortunately.

Michael Finocchiaro

Sorry, Brian. It's okay. Um, well, it's been fantastic and um, I'm I'm looking forward to uh to talking the next time uh, when you guys have some big stuff to announce. And um, uh, thank you to the audience for listening. U, please feel free to contact Jonathan and and go to the website. And, uh, we'll I'm not sure how many more of these are going to do before the big summer break, so we might be seeing you September. Um, in any case, thank you, Jonathan. Say hi to the whole team and, uh, and we'll we'll talk to everybody soon. Have a great summer.

Jonathan Girroir

It's good. Thanks, guys.

Michael Finocchiaro

Uh