199: ‘Autodesk’s Vision for Architects’, with Amy Bunszel

199: ‘Autodesk’s Vision for Architects’, with Amy Bunszel

Evan Troxel: Welcome to the TRXL Podcast. I'm Evan Troxel, and in this episode I'm talking with Amy Bunszel. Amy is the Senior Vice president of AEC Design Solutions at Autodesk. Her unique path, starting as an electrical engineer in the semiconductor industry, founded a 3D wire harness design startup, and eventually leading teams responsible for products like AutoCAD and Forma has given her a rare perspective on both the pace and the depth of change in the AEC tech landscape.

While I was in Boston in June at the 2025 AIA Conference on Architecture, Amy and I had the opportunity to sit down and talk about how Autodesk is rethinking design workflows and interfaces from the ground up.

We explore the evolution of tool usage and customer expectations, the power of product analytics for surfacing better ways to work, and how AI is becoming a design assistant rather than an opaque replacement for creativity. Amy shares insight into how Autodesk is creating tools for the early stages of design in the Forma platform using context aware data, to help architects make smarter decisions sooner. We also discuss collaboration in the cloud, the future of AI assistants like the Autodesk Assistant, and why doing the dishes with AI is a better place to start than chasing hype driven use cases.

A key theme from this conversation that we're seeing a lot these days on the podcast is the move from a blank screen to a rich contextual starting point, one that reflects real project goals, real sites, and real outcomes. We dig into how that shift powered by tools like Forma and intelligent data integration, like the newly announced GIS integration, which comes through their partnership with ESRI, stands to reduce upfront rework and change how projects begin.

As usual, you can find links and other resources in the show notes, which are in your podcast app if you're a paid supporter of TRXL, and if you're a free member or casual listener, you can find them at the website, which is TRXL.co. Sign up on the site while you're there so we can be in contact.

And lastly, you can really help the podcast by sharing these episodes with your colleagues. So now without further ado, I bring you my conversation with Amy Bunszel, Well, I would love to hear kind of your story, 'cause I don't think I've ever heard it like, like your, and, and as long as, as you want, as you, but, but like how have you kind of worked your way through the industry?

Amy Bunszel: That's a great, that's a great question. So, and, and I was expecting it 'cause I know you always ask, this is pretty much

Evan Troxel: how I start.

Yeah,

Amy Bunszel: no, it's great. Um. So I am an electrical engineer by background, not, um, not the MEP electrical, but more of the, you know, machinery, electrical. I worked in semiconductors for a few years. Yeah. Interesting. And then with some, uh, colleagues, we started a company doing three dimensional wire harness.

Design. Okay. So if you think about all the wires and cables that would go in an MRI machine, a piece of factory equipment or construction equipment, you know, in 1996, this was a problem that was unsolved. And so we ran that company for about eight or nine years. And then in 2003 we sold the company to Autodesk.

Evan Troxel: Oh, okay. Yeah.

Amy Bunszel: So I came, I moved from Boston to Portland, Oregon, and I. We started building parts of Inventor and so we moved that functionality into Inventor. It's still shipping today. Yeah. So I'm super excited that, you know, it endured, uh, for 20, 20 plus years. And about eight years into my time at Autodesk, the AutoCAD leadership position was open.

Okay. And I thought, Hmm, that would be interesting. You know, and I, speaking

Evan Troxel: of things that are still going Yes. And

Amy Bunszel: a gift that keeps on giving, that's absolutely AutoCAD Uhhuh. But you know what a fascinating. History. What an interesting time for AutoCAD. As you know, BIM was coming into its own, you know, 3D parametric mechanical modeling had been around for a long time.

Like what was the role of AutoCAD going to be for the future? So I uprooted my family and we moved to California and it's, that was really my first foray in to AEC. Wow. And so it was, it was great trying to figure out the, the future of AutoCAD in a world where, you know, there was lots of other.

Different kinds of things happening. And we really decided to go back to the roots of AutoCAD and focus on on 2D, drafting, documentation, drawings, all the things that people had really been able to get value from. And now over the years we've got it running on desktop, web, mobile. Mm-hmm. It runs, um, it runs AI and all kinds of great things.

So it's an extremely modern version of AutoCAD. Yeah. And then I ended up. Uh, over time, uh, moving more into the AEC solutions. And so it's been a lot of fun.

Evan Troxel: Wow. And, and so you've seen a lot change.

Amy Bunszel: I have over that time. Yes.

Evan Troxel: And being at a technology company like Drive, trying to drive change. Yes. Right.

Uh, I'm just curious, like what do you think like the big hits in your tenure have been? I mean, obviously BIM is a big one, but Yeah. But like, like as far as like things that have actually moved the needle for architects and engineers and contractors, what would you put it kind of your top of your list?

Amy Bunszel: Wow. Well, I would, I'll have to go start with the internet really. And thinking about the, the, you know, there was a, a, a joke, I guess way back in the early, the late nineties where, you know, everything needed a.com and you couldn't, yeah. Couldn't raise money or do anything unless you created some sort of, you know, internet based application.

But, you know, in the end. That's really where we are, where we are today. So I think the adoption of, of the internet and then thinking about connectivity and, and sharing, it was a long, slow start until we got really to the cloud. Yeah. And then the cloud happened much faster. I think, thanks to COVID. We had customers who were, you know, over my dead body will be put.

Any information in anybody's cloud to, oh my gosh, it's, it's COVID. We're gonna implement cloud capabilities in the next like week and a half, as opposed to having a six or nine month change management program. So those two for me, but of course now it's, it's happening even faster now with ai. Right? Every morning I wake up and there's some new development and ai.

Evan Troxel: I was at the keynote earlier and, uh, there are many things that you just said that I want to tap into, but that she, that. The keynote speaker showed the graph of like adoption of to a hundred million users. And you know, it was like the internet was seven years. Mm-hmm. And all of these things all the way down and.

You know, like chat GPT, it was like, it was a very short period of time. Yes. Right. It was like two years, right? Or not, not, it wasn't two years. It was, it was a matter of like months. Months, yeah. Basically from when they, when they launched and, and so yeah, like that speed of adoption. And I've seen other graphs like that in the past, like, like plain old telephone system at pots, right?

Mm-hmm. And then it was like, because not everybody had a telephone and that was a very long time period and everything keeps getting shorter and shorter and shorter, and it's like. How much can we actually take? Right? Like real people?

Amy Bunszel: Yes.

Evan Troxel: Have to and, and, and yet like, um, because of all these other underlying technologies, the phones in your pocket, like maybe the, you know, the EarPods, like, there's so many things that are so interconnected that it does make it easier.

And the interface matters. Mm-hmm. So much more than it used to. Right. And, and I think because it's interesting when you, when you sell a bunch of products that ha probably have a two to three year learning curve mm-hmm. Like to really get into it, not just surface level stuff. In a world that is, like you say, you wake up every day and there's a new, some new developments with ai.

And being kind of bound at some level to those frameworks that you've developed over decades. Mm-hmm. How do you think about your customers and how you can move to react to their needs so that those adoption curves happen so much faster because they are.

Amy Bunszel: I think about it in two different ways. So I think about our install base using those products where they've invested, you know, a career Yeah.

Worth of, of learning how to use them. And even then maybe only tapping into 25 or 30%. Yeah. Of what, what's possible. Or maybe

Evan Troxel: only needing 25% of what they offer. Yeah.

Amy Bunszel: Or in some cases there's five different ways to do something. What's the most optimal way for me? So one of the things I'm happy about is probably about five years ago we started investing in understanding usage analytics, and then being able to.

Tell our customers, oh, hey, there might be a better way to do this. Or We've sensed you're getting frustrated because you keep, you're sort of in a loop of doing, doing some activity. And so using their data and their usage information to kind of feed back to them, um, guidance. I think that's been really helpful.

And we're seeing. Very good uptake on people exploring new capabilities or doing things differently and getting, getting unstuck. So how

Evan Troxel: do you think, how are they finding those things? Because like a lot of times, like, oh, I didn't even, I don't know what I don't know.

Amy Bunszel: Yes. And

Evan Troxel: so are you putting stuff in front of them to prompt that?

Amy Bunszel: We are doing a few things. Yeah. Uh, we, we started very gently with email. Um, now we put things in some, now we have more portals and other ways for people to get it. Yeah. So many ways. And the best way is really to surface it in product, but to do it in a way that isn't distracting from their, their work.

'cause you don't wanna be clippy like back in the day where it's so much, it comes

Evan Troxel: up so much nowadays, it's like, oh, I just wish I, that this thing was watching me on some level. Mm-hmm. And other people are like, no, of course I don't want to.

Amy Bunszel: Yeah.

Evan Troxel: Watching. But there's, there are people who want it to be proactive.

Yes. In nature. And I think that's really interesting that we've come back to that and I'm just hearing it. Mm-hmm. A lot from people. It's like, oh, I, I wish I had a, a recommendation engine

Amy Bunszel: Yes.

Evan Troxel: In the product. Or I wish I had a thing that would say, you know, there's a better way mm-hmm. To do this.

Amy Bunszel: We, we agree.

And also it gives us ways to surface new features too. And to find those frustration points and then. Work on fixing them. So that's kind of how I think about the, the, the core portfolio that people have been using for a very long time. But with new things mm-hmm. We cannot create the burden of learning that we have with the, with the current tools.

Yeah. So, and that again is about you being smarter, thinking differently about how we're building the interfaces, thinking differently about how people learn. Think. I mean, I remember back in the day we had manuals. And we got rid of the manuals, like the decision process to decide not to print, you know, an AutoCAD or a Revit manual.

A dead document was a big deal. And now nobody wants to really learn that way, right? They wanna learn in bite-size pieces. They wanna in the moment learning when they're actually trying to do the thing. So we need to use technology to make that all a lot easier for people to onboard to the software.

Evan Troxel: Former educator too. Uh, and, and so spending a semester or a quarter teaching people how to use a product to some nu percentage. Mm-hmm. Like you mentioned, 25%. Yeah. Yeah. Maybe. Right. Um, and, and going through the book, through Paul Bins, tutorials, for example. Right. Like chapter by chapter. Um, it's interesting to think about how much, how learning has shifted.

Mm-hmm. Right? Like, it's like even. You know, I run up against a problem in a product and I'm gonna go on Google and look like my wife. Uh, you know, she, she's, uh, also does BIM and drafting and all these things, and she's just like, uh, I can't find the answer. Right. I just want to find the answer. Yeah. And so it's like, okay, search, search, search, dead, end search, search, search, dead end.

And it's like, yeah, it kind of needs to be right there mm-hmm. Right at the surface so that we don't have to go digging and, and ultimately find a dead end. And I think about the depth of the products and, and the breadth of them. You're serving a lot of industries, so does this mean like more specific products for specific things rather than products that do so many things, do you think?

Or like when it comes to road mapping this stuff out, how do you think about that?

Amy Bunszel: I think it's more. Different experiences based on the workflow the customer's trying to do. You know, we've tried over the years to understand the different personas of our customers, but there's many, many, many, there's so many, many.

So, but if we think about like, what are you trying to do? Like if you're trying to do, you know, a carbon analysis, maybe you should just get a pallet of tools that, that help you do that. Yeah. Um, or if, if we know you're doing. You know, detailed design, you're gonna get, you're gonna get that pallet of capability as opposed to if you're doing drafting or if you're doing different things.

So for me it's about not overloading the interfaces, but being flexible, but also being more intuitive about what the customer's trying to do and then surface up the capabilities are gonna be needed for, for that, that task.

Evan Troxel: Yeah, I think about like, uh. One of the vendors here, parallax team. Mm-hmm. Aaron Mahler been on the podcast before and he makes templates for Revit.

Right. And they have everything in them. And, and his, his thinking, at least as far as I understand it is, and he can correct us if we're wrong, but it's like start big and then throw away everything you don't need so that it's all encompassing, right? Mm-hmm. And, and I think of like modules inside of applications like.

Like an animation package might have like a, a materials module and then an animation module, and then a rendering module, and it's like, it, it just kind of filters out the tools that you don't maybe need while the, obviously it's all one overall encompassing platform. There's so many ways to do it right, and you guys have so many tools in the portfolio.

Is there any way to even bring consistency to, and not that it even needs to be, I mean, different, there's different minds, right? And they think differently. There's, there's so much to think about. We use

Amy Bunszel: the word, uh, coherence more. Okay. Because I, I think consistency for the sake of consistency brings you to a least common denominator kind of place.

Mm-hmm. But software should feel like it, it. Was meant to be together if it's part of a, a family of, of, uh, capabilities. So I think, and, and don't be different for the sake of being different, right? If zooming in, zooming should work in the same direction in all of the, all the capabilities, so you have that muscle memory and you can focus on learning that

Evan Troxel: intuitive,

Amy Bunszel: the new things.

But, but for us it's also like to your point about the, the big everything universe will. It, maybe you can tell me I'm working on, you know, the multifamily residential project and then magically, you know, all those things. We don't think you need go away. Obviously you can go back and get them. Yeah. But we can, we can be smarter about how to start and, and that's interesting to me about our new products.

Particularly with form is we start with your location and the context and all the data you wanna gather about the project so you can start to make, make those good decisions as opposed to the very intimidating kind of blank screen that we all used to start out with back in the day. Yeah,

Evan Troxel: I mean it uh, that's something that the keynote speaker also mentioned.

She's, I hate starting with a blank screen and that I thought that was interesting and insightful because. Like, we as architects have always started with the blank screen or a blank sheet mm-hmm. Or whatever. And I know that this has kind of come up, especially in the last five years, let's say. It's like, why do we always start with and, and just kind of challenging that base assumption that that's where design starts.

Mm. I think that's super interesting. Right. And so, and to your point with form, right, like starting with context, like every project has a place mm-hmm. On the earth. Yep. And, and yet there's still so much design that happens like right here on this like cloth, right, there's table cloth. I think that's super interesting to, and it, and it, and it's hard is how, talk about maybe the, the mindset shift in the designer.

When you talk about something like that, do people say like, what, that's not my process, or. Or, or you see it as it's going to be your process. I'm just curious because there's so many different approaches as always.

Amy Bunszel: You know, I haven't heard anyone say, that's not my process. Even if, even though I know it's not, there's certainly change management involved here.

What I hear people get excited about is delivering better outcomes on the project and, and being able to have that information upfront. So it's not like we're getting towards the end and we need to go double check that we hit, you know, certain goals for the project, and then we have to go back and rework, or we have to go ask for more money or we have to do value engineering or something.

So I, I hear optimism that. Better information will provide them the ability to do even, even better work and to focus their efforts on the places where, where they can make a difference and to hit those, you know, key project milestones. Yeah. The other thing I would love to see us in the future is if we could take the spec.

Or the brief or whatever you start out with, and start to kind of transform that into the context and into the objectives and the goals at the very beginning of the process. And then help track those throughout the entire Yeah. Process as things evolve.

Evan Troxel: Do you, so yeah, maybe like what are you thinking in those regards?

Like, 'cause I, I haven't heard that before, but it totally makes sense, right? It's like infusing the different inputs Yeah. Into one location. Mm-hmm.

Amy Bunszel: Kind of a thing. Kind of, and, and thinking about, you know, the, some, some of the outcomes are, you know, very tangible and measurable. Some of them are mo a little, a little less, but to the, to the extent that you can let the software know where you know who you're building for.

Yeah. Or what you're building. Um, I went to the keynote yesterday in the tap, um, co conference and it was all about people. Like, we're doing things for people, right? They have outcomes and impacts on the, on the environment and everything else, but we really wanna be thinking about the experience for the people.

Yeah. So the extent that we can learn from that, maybe compare it to prior projects that had similar goals and understand what was successful in those projects. All that stuff should be at our fingertips as we're starting new work.

Evan Troxel: Yeah. Well maybe you mentioned format, like, and that's obviously the. A lot of energy mm-hmm.

Is being put there. And I've had the opportunity to interview Carl on the podcast and, um, I love talking to Carl. So smart. He's so smart. Uh, and, and I love kind of his ability to, I mean, we've seen a lot of cool things happen. Mm-hmm. I think with, with that product in particular and kind of seeing the next gen of.

Whatever you wanna call this a platform. Yeah. Or, or whatever. But, um, I'm, could you just kind of give a little bit of a, like what's happened over the last couple years since form kind of officially launched as for Absolutely.

Amy Bunszel: Yeah. So that was May of 23, I think. And I remember you talking to Carl. Carl, Carl.

As a founder was surprised that it was so hard to do some of these things, right? To bring the data together, to have that great experience in that early phase of, of conceptual design. So, you know, we've been out in the market now for about 18 months or so, and. And we've continued to, you know, add the great thing about a web, you know, web product like that is you can update fairly frequently.

Yeah. So still really focused on kind of site planning, but lots of, lots of additional information about context. The thing we announced this week was our ESRI integration, which people are really, really happy about that. 'cause there's, they're using that data across their ecosystem anyway and to bring all that information upfront as they're making decisions about the project have been, have been really critical.

Uh, there's also, um, our new. Much deeper integration with Construction Cloud and, and Autodesk Docs. So now when you create a project, you can, you can create a project using the form of data, using all the same permissions and controls that you're using across a CC. So we're really bringing, you know, design and make together in this common project ecosystem Okay.

Across the life cycle, which is, is great. So it's so much easier now for people to kind of get in there and, and start these, these projects. And then the third. The third one is collaboration. Everybody in the industry is talking about collaboration and in many different forms. And so we have the former board, which has been out for a few months now, and that allows you to have, you know, many different ways of collecting all of the, the options and variants and, and things you've been working on, on a project in a way that you can collaborate with people who don't have to be experts in, in the tools.

Evan Troxel: So boards is kind of like a, a Miro board, or it's a, it's a whiteboard. You can put images, comment, it's kind of an infinite canvas.

Amy Bunszel: And you and, but you can also get in and like, have people collaborating with their presence. Right. You know, their little avatars or arrows or however, from remote

Evan Troxel: locations.

Mm-hmm. All in the same file at the same time and, and contributing.

Amy Bunszel: Yeah. I learned that I guess during COVID a lot of. Architects started using those products as, um, ways to kind of share Yeah. Share their designs and so it's certainly caught on. And now if we can do it with set of screenshots, right? This is live information that is current and up to date, and, and you can easily add to

Evan Troxel: kind of view ports into models mm-hmm.

If you wanted Yeah. Yeah. Talk about that Esri integration, because I think a lot of architects don't think of GIS. Mm-hmm. When it come, like in the power of what's in GIS. And so talk about why you decided to, to include that at this like early stages of design, right? Like I guess that's where what we're talking about with form and making better decisions earlier.

Mm-hmm. Um, and, and like what, what is the secret or the magic sauce of, of GIS data so that architects can understand why they might be interested in it?

Amy Bunszel: You know, a lot of it is, is. Context, right? And thinking about, and, and with Esri context, it's, it's deep, right? There's just tons and tons of data and information in there that, that help you with, you know, if you're thinking about, you know, zoning and, and all sorts of different types of data that can be overlaid in sort of a map environment.

So it's, it's, it's pretty incredible in terms of all the site planning that you might do. And then when you layer on top of that BIM data, now you've got the, the full picture of kind of the, the. Horizontal and vertical situation that, that you're dealing with. And the, the, the cool thing about Esri also is if we think about working with municipalities, right?

Esri's really embedded into most state and local and, and the federal government here in the United States. So there's really rich data sources that you can pull from there as well to leverage.

Evan Troxel: I remember talking to somebody and they were like, why don't we, and this was before like. You know, civil engineers have been using that kind of data for a long time, but it was like, why don't we have that?

When I was running digital practice at our firm, it was like, we should be looking at GIS because then we're talking about even like demographic information. Yes, right. Tying into census data. That's part of that GIS. Yeah. I mean, I think and like thinking about just additional context that is otherwise really difficult to, to.

Locate on a map. Mm-hmm. Right. I think getting back to that idea of a project is, is going to sit somewhere. Yes. And that somewhere has this bubble, right? Yes. And, and, and what's in that bubble? A lot more than we typically would apply that solar. I can think about

Amy Bunszel: crosswalks. Bus stops all traffic.

Evan Troxel: Yeah. All this data

Amy Bunszel: that, that would impact maybe how you situate something on a site ab

Evan Troxel: Yeah.

Amy Bunszel: Um, and give you, again, bigger, better experience for, for the people you're building for.

Evan Troxel: And so some of the other, um, layers. Are noise, right? Mm-hmm. Uh, and sun exposure, things like that. Yeah. Is, are what else is on that list?

Amy Bunszel: Yeah, so we do a lot of our, um, analysis also, uh, is, is trying to leverage kind of the, the environment.

And so everything from, we do noise, we'll do sight lines, we'll do, um, we, you can find, uh, heat islands or, you know, if you're building a, a campus, there's too much shade in some areas, not enough shade in another area. All those things you might not figure out until. You know, way too late in a project to have an impact.

And you can make those changes upfront and then collaborate with your stakeholders, whether it's, you know, the, the, the owner or if it's the, the community, whoever needs to be involved in that collaboration.

Evan Troxel: So now shifting gears a little bit with the, the docs integration with form, can you just talk about like what obviously that that matters, especially at an earlier state and you talk about permissions and access and all of those kinds of things on a project.

And this is a web-based tool, right? Mm-hmm. So people can play in the sandbox from anywhere. Can you talk about like why, maybe where you're going with that and why that's important? Now,

Amy Bunszel: well, docs is our, uh, common data environment for our, all of our a a C products. So all of our solutions, whether they're desktop or cloud, are starting to integrate into docs.

And so that gives you that common data environment so that you, you have a source of truth that you, you can understand and it's easier to. Share the data across phases. One of the other great things about Docs is that we have a bridge capability where you can invite other people. So you know, you might be the architect, someone else might be doing the engineering.

You can, so

Evan Troxel: outside your organization,

Amy Bunszel: yes you can bring them into the project. They bring the data, but they still own the data and they can still control the data. So we're, again, that's another type of collaboration is how do we, instead of throwing things over the wall, how do we bring the data together and create these aggregated models that we can work together on.

Evan Troxel: And are you able to leverage that? Okay. So like thinking about AI and LLMs and, and like the way you can leverage those types of tools to, I'm trying to just want, I'm, I'm curious like, how are you thinking about all that because things are changing so quickly. Yes. What can we do? Well, certainly

Amy Bunszel: having, having data in a.

Commonplace gives you better, is a critical way to start trying to figure this stuff out. And then it's about understanding what, what problem you're trying to solve. 'cause you can have a lot of data, but you really have to understand do you have the right data for the use case? Yeah. So one of the things we did in Construction Cloud was an ask your specification capability using an an LLM.

And so this, this is great. So you've got a thousand page document instead of having a poor human being in a. Highlighter try to find things. You can, you can query for where are all the glazing requirements or where are the concrete specifications, and we annotate it with where it came from. So you can go and citations, check in the document and everything.

And so that's really just the beginning of querying all of that data. Yeah. Um, in a, in an intelligent way. And then from there we can start to think about insights. So what kind of, what can we derive from that data and then let you know, um, risk is an area, certainly in, in construction, that everybody is.

Interested in whether it's cost, risk, schedule, risk, just interested, risk. Risk, right? I mean, you can blow a whole project with a, with a, with some of those things. So, so we already have some capability that will look at, um, invoices and project data and look for patterns and maybe say, oh, this, every time you use this vendor, you have this problem.

And so helping the humans just have more information at their fingertips to make, make decisions.

Evan Troxel: So this is where I, I'm gonna, this is what I do on this podcast. I bring it back. Yeah. And I think about clippy. I think about clip, not clippy. Iss not the thing I always talk about, but it's like in our story today, like this recommendation, insight, mm-hmm.

Thing, whatever you want to call it. Right? And, and like, yeah, your contracts in there and your specs are in there. And obviously your models in there and your and and analyses and all of these things. And, and like as an architect, I just want, obviously there's like, we're limited by our experiences. We're limited by.

Who's on our team. Mm-hmm. Potentially. Um, and obviously that's where a lot of magic happens. But this, you know, the thing that I, I think about with AI is like this other teammate Yes. Who, who can keep all of that stuff in, in mind. Like, maybe not in mind, but like, I just think of like, oh, oh, you're, you're, you're one, you know, one counselor office short.

It's like, it's like this. This just thing that's running and checking to see if you're meeting the brief, if you're meeting the contract, if you're meeting the timeline, the schedule. All of those things, the meeting minutes, right? It's like, yeah, you

Amy Bunszel: followed all the steps. Right? Did you, because this is all, all part of

Evan Troxel: the contract documents, right?

Mm-hmm. Like, and, and not everybody, like you said, the poor human with the highlighter, right? Seriously. Yes. Like and a

Amy Bunszel: well-educated, poor human with the highlighter. Yeah. No, yeah. It's who probably doesn't wanna be, who would like to do more creative things

Evan Troxel: to, yeah. And, and, and, but it's like, it's like this, this system, I would, I would hope is kind of, you know, feel like, feels like a teammate.

Like, yes. Oh, thank God that was. Caught that. Yes. Or recommended that what a cool, and, and so like when it comes to like those kinds of potential things Yeah. What do you, are you excited about that? Like, is it that, 'cause that to me seems exciting as a possibility. Mm-hmm. Right? For, for these things to be.

Again, clippy like, Hey, it looks like you're writing a letter. Would you like, right. I I I just feel like the potential's there now,

Amy Bunszel: or an additional collaborator. Yeah. In a way, maybe the, maybe the, um, the knowledge base of, of something that you can then query as part of a collaboration with other humans and with, with, um, you know, AI agents.

And then the, for us, what's also exciting is it's not just about the insight, it's about can you take the action? So does it, in the beginning. I'm, I'm sure trust will need to be built up. Right. But over time, if the, if the AI is really good at telling you about the issue and then they can take action to resolve the issue.

Yeah. And you, you build up this paradigm of trust or, uh, then you know, you're, people are gonna be a lot more productive and be able to focus on the, you know, the fun things that they wanna develop for people.

Evan Troxel: Oh, so you guys announced Autodesk intelligence at the mm-hmm. At au and, and au the message. Came, I think for architects it was like, I, I was unclear still what does that mean?

And obviously I, there was an emphasis on low hanging fruit and value, right? Yes. And, and I even more so all the time, like obviously there's a lot of hype around AI and this amazing big cool and it's like literally we just needed to do like the. The thing like we just talked about with the specs and linking what's there to what's in the 3D model and saying is this the right thing?

Is it the wrong thing? Like those tiny things would be so valuable for architects. Yeah. And so are you still really like beating the drum of like fundamental, valuable level?

Amy Bunszel: We are, and we tend to call them productivity features. They're things that just, you know, automate a workflow. Um. Check for errors.

Those kinds of things are where we're spending most of our time right now. And

Evan Troxel: how are customers responding to that? Because like, I, and I'll preface this by saying like, a lot of people like, like the shiny object syndrome, but I think as coming as a formal former digital practice leader, it's like people say AI and, and, and we're saying, we, as in digital practice leaders we're saying like, well, what do you mean?

Like what? Like, because they don't even know what they're saying. They're right. And it's like, I, it sounds like you're saying. Data. It sounds like you're saying, but but you're saying the word ai, right? And it's, and it's somehow magical. Yeah. And it's not, right. It's like we have to be really proactive about how and when we, we employ these tools to do these things, but, but they have, like, we really want them to do the stuff down here, like not doing the art so that we can do the dishes.

Example, right? Like, we wanna do the art and not the dishes. So I want the AI to do the dishes. Is that what we're talking about? It

Amy Bunszel: is. And, and, it varies though, 'cause some of the newer tools, like the, the AI we have in Forma is doing, uh, we're training on surrogate models so we can do near real time predictive analysis.

And so that's a great. Use case for what kinds

Evan Troxel: of things? Noise. Okay. Some of

Amy Bunszel: those things. So you can do a full blown analysis, but we can also do it so like you can iterate really quickly. Mm-hmm. And then when you're ready, you can do the, the a more detailed, bigger, more detailed one. Detailed. Okay. So those kinds of things allow sort of the best of both worlds, like.

Rapid iteration using a good enough answer that you're getting through the, the AI training. And then you can go and do the more, the more detailed analysis. And so that's definitely a partner like AI as a partner in your workflow. And then the other, a lot of the other productivity things are just like routine, mundane tasks.

We've delivered a few more of these on the manufacturing side to date. Uh, for example, we can, uh, automatically constrain drawings. So if we, we can look at something and tell. This, most likely these are supposed to be perpendicular, and this is supposed to be parallel, and we can start to put all the constraints on there.

And so those kinds of things will start filtering their way into the AEC. Workflows And I'd say most of our customers are very optimistic and willing to, to try things. Although we did see in our designer make report recently that I think they're getting a little, um. Um, there's still about 68% of them are still very excited, but that's 11 per points or so, less than it was the last time.

Oh, interesting. We did it. So I, but I think people are realizing this is a little bit complicated and I have to be thoughtful about how I'm gonna deploy these things across my whole enterprise. It's not just the Autodesk piece, it's the whole business that can benefit from

Evan Troxel: Well, they're getting bombarded with

Amy Bunszel: Yes.

Evan Troxel: And so I, I'm curious, like, like when I asked the question about like, how are they responding? Are they responding like, are Autodesk doesn't move fast enough, or are they, are they actually like really cheerleading for you to focus on the stuff that we were just talking about, which is like that, the mundane stuff.

Yeah. Right. Because that to me is where we, you mentioned it earlier, right? Like, we want to spend our time doing more valuable or better things, right? More exciting things. Yes. Uh, and so like in order to get there. I think we need to start here. That, that makes sense to me as a strategy, but are you still hearing, like how come you don't, like, where's the big flashy stuff, Autodesk?

I, you know, I, I'm just curious at what you're hearing.

Amy Bunszel: I'm hearing more about the, we want more productivity features faster because that's what they're experiencing with some of the other vendors, like with the Microsoft where it kind of came in and was kind of pervasive and Yeah. And now they're.

They're trying it all out. So they wanna know, you know, where's a copilot for Yeah. Every single one of the, the products that you have. And so we're being thoughtful. Um, we're not just thinking about it from a, a Autodesk assistant would be our equivalent of those types of, of workflows, but we're also thinking about features and capabilities that will be driven by, by the ai.

Evan Troxel: Yeah. Well, it's exciting future for sure. And, uh, and I. I mean, is there anything else that you wanna talk about with form that, that you're coming out with or anything, any other subject as far as a i a architects are concerned.

Amy Bunszel: So what, what goes hand in hand with the, the, I think the collaboration conversation, the AI conversation is data.

Mm-hmm. And so we've been working for the last two years now on this AEC data model, which is right this granular. Data model that is much more accessible. It's it's stored in docs, our repository and then you can use an an API to get at that data versus in the past where you had to fire up Revit, use the Revit API and it was a much more challenging right process.

So I think that's going to help with collaboration and that's gonna help people feed downstream processes, get more value out of all this data, derive other insights. By matching that data up with their ERP information. So we're definitely seeing a, a while people want automation in their, you know, design workflow.

They also want that data to percolate throughout their ecosystem better and be able to use it for more insights in other ways.

Evan Troxel: How does that translate to medium and smaller firms? Right. Who are like, what does that, I don't know what API means, I would know, I wouldn't know how to access granular data.

Mm-hmm. Uh, how, how does that translate to them? Can it, I guess? Well,

Amy Bunszel: we we're doing a few things to, to lower the barrier there. One is we built a whole family of connectors. So like if you wanna connect your data to Power BI or some of the dashboarding tools that come from Microsoft, we're making those sort of point to point integrations easy as well.

So you don't have to be an expert for all of that. I

Evan Troxel: think one of the things that kind of blew me away today and, 'cause I haven't done this myself, I haven't thought of a use case for myself, but, uh, and, and the keynote she talked about. Um, you know, she went to Claude and she just was like, I'm gonna be interviewing and, and I don't, it is just a fictitious set, set up, right?

But it's like, I'm gonna be interviewing somebody for a position in five minutes and I need a way to evaluate them. And it was like, build me a tool that will do, and it like on the fly, can actually just build a quick little app that then allows. Assign a rating and it, and it, it was like, here's the main questions that you should ask this candidate based on the role that they're, and I just thought like, it's actually really, really incredible to be able to just kind of prompt your way mm-hmm.

Into making something useful very quickly. And I kind of, you know, connecting the dots of connecting into granular data. It seems like there's a way potentially in the future where it's like. I need to change, you know, all these doors from a, from a this to a that. Yes, I need to change all the frames from a this to a yes to a that through a very intuitive interface.

Yes. And then just say like, okay, now how do I check this in, make sure it actually happened. Mm-hmm. Right. Giving kind of a checks and balances to the system. But it seems like you are kind of positioning. Your platform to be in a place where people could be able to do things like that with it. Yes. Yes.

And that sounds pretty exciting.

Amy Bunszel: Yeah. And we would put that under the Autodesk assistant umbrella. Mm-hmm. Thinking about creating plugins for all the products where, where you could ask. Revit to change all these doors to this, and then you could, then you can ask it to report all of your door types. And so you can, you can have, it happens in software development where the coding assistants are both writing the code and writing the test to check the code.

Evan Troxel: Right.

Amy Bunszel: So you're kind of getting the best of both worlds. You getting the

Evan Troxel: offer right, and build the dynamo script and then have to worry about how you wire it up and running that and sending it to a friend or no, it's just like, no,

Amy Bunszel: and we can connect to the, we can connect the, the. Prompts to the Dynamo APIs and you know, a lot of these things will happen more naturally.

It's pretty cool, naturally. Cool. And it's also an interesting way to learn because, you know, you can, I know in on the coder side of things, like people are, you know, learning on the fly with the AI as a partner as opposed to do know just giving, giving it, we're not giving it all up. Yeah. To the ai. Yeah.

But that, that's where I think people who are curious and willing to try these things out are gonna. Quickly help us expand the use cases because I, I feel like every, almost every day something else happens where I'm like, oh, well now maybe we could do this, or Yeah, maybe we can connect to that. And I dunno how far it's gonna go, but it's gonna change a lot of things.

Evan Troxel: It's gonna change a lot. Yeah. Well, thank you so much for taking the time to talk about this today.

Amy Bunszel: My pleasure. Pleasure

Evan Troxel: to meet you.

Amy Bunszel: Wonderful. Thank you.