In Stay Sharp Episode 143, Build vs. Buy in the Age of AI-Assisted Coding, co-hosts Juliann Grant and Jonathan Scott sit down with returning guest Jonathan Girroir, Technical Evangelist at Tech Soft 3D, a software company that creates the building blocks for digital engineering. In episode 139, Digital Product Series – Tech Soft 3D, Girroir talked about some of the emerging trends in digital engineering, most of which center on what AI brings to the table. This time, he’s breaking down why the idea of writing your own PLM software might not be so farfetched, how far vibe-coding can take you, and what to consider when doing anything outside of your manufacturing plant with your data.
Who’s Building Their Own?
Jonathan Girroir, Technical Evangelist at Tech Soft 3D
Girroir explains that there’s a lot of growth happening in a couple areas of the industry because new AI tools have lowered barrier for entry. “The foundational pieces are market-hardened, commercially available tools,” he says. “All you’re doing is adding on top of it the decoration, the user interface, the glue between the different pieces.” The number of startups creating their own software tools has exploded in the last 12 to 18 months because they can quickly build a prototype, pitch it, secure funding, and get it to market. The speed is partly to do with the availability of the kind of powerful components that Tech Soft provides, including data translation graphics, simulation, and more. But, Girroir adds, “The AI agents and copilots and vibe-coding tools allow you to take those with documented APIs and connect them together, and then add on top of that your special sauce.”
The other market segment where DIY is booming isn’t what you might expect: large organizations with deep expertise in CAD and engineering. While their size means experience, entrenched software, and resources, they also may have a lot of specific needs that aren’t being met by what’s available on the market. It’s now easy for them to build the tool they’ve always wanted. Even within some of the larger independent software vendors (ISVs), Girroir says, “Their R&D teams are able to pitch ideas and build real products in R&D instead of putting together slideware. In a day or two or a week, a software engineer is able to actually show people, ‘This is what we want to do. This is where we want to go.’”
The best advice Tech Soft offers is don’t reinvent the wheel—don’t build your own platform just because you want to build something. “Lean into where your expertise is,” Girroir says, noting that Tech Soft is the expert in data translation, graphic, simulation and analysis. He adds, “Let our experts be your experts, and focus on how you can distinguish yourself within your area of expertise and build out those workflows. Honestly today, data translation is a commodity. You can get it from Tech Soft; you can get it from other vendors. You’re not going to distinguish yourself on top of that.”
Data and Delivery
Girroir says the future of the data model is APIs. That means hosting your data somewhere and not worrying about its format for integration. “Instead of working through a file-based structure, you work through a series of application interfaces where you’re actually requesting data in a particular format,” he explains. “That is where the ultimate flexibility is, where you have whatever data on your server or locally and you’re able to request from that, ‘Give me X, Y, and Z.’” Jonathan loves that idea because you can stop worrying about how your data is saved. “Worry about how you interface with it, how you ask for it,” he says.
Storing data on the cloud—as all the new startups are doing—will bring plenty of advantages, as well as some challenges. Being able to request only the data you want instead of full files or packages will speed up load and interaction time, and the potentially infinite compute power available means complex transactions like simulations could get faster. Between the cloud and API access, you’ll need to carefully consider your data ownership and security so that you’re only sharing what must be shared. “In some ways you’re getting more security and ownership of your data by allowing this very granular exposure of what you want out there,” Girroir says. “With great power comes great responsibility—and maybe security risks.”
Learn More About Where Digital Engineering is Headed
The full podcast contains more discussion and details on a variety of topics, including how maintenance costs need to be first defined and then considered in the buy vs. build evaluation, why Girroir doesn’t think the industry has developed “the model” in MBE yet, the data sovereignty questions you’ll want to consider, and more.
Be sure to check out Stay Sharp Episode 143: Build vs. Buy in the Age of AI-Assisted Coding and join us each week for a new podcast.
Follow the Razorleaf Podcast, Stay Sharp in Digital Engineering on:
