From Innovation to Transformation: Building an AI-Native Practice with Evelyn Lee

A Welcome to Practice Disrupted, a podcast exploring how architects can build better businesses, embrace emerging technology, and create more sustainable, fulfilling practices. I'm your host Evelyn Lee, architect, angel investor, and founder of Practice of Architecture. Whether you're running a firm, starting your career, or reimagining what's possible, this podcast challenges you to think differently about the future of architecture. When I was practicing, finding a detail meant digging through old models. That's why I'm glad to have Piros sponsoring this episode. They make every detail, family and standard searchable across your firm and ensure that nothing gets used blindly. See it@piros.com that's P I R R O S.com. Hello disruptors.

A Welcome to this week's episode of Practice Disrupted. This one runs differently. I am in the hot seat and the questions are not coming from a guest. They are coming from Sage, my AI agent. I wanted to stay curious about what this technology can actually do, so I turned it on myself and let Sage run the interview. A few honest notes first. I came in with a set list of questions, but Sage was instructed to probe deeper whenever she felt it was warranted. So a lot of what you will hear are her follow ups, not my script.

A I'm only able to record in 20 minute intervals, so this is five conversations spliced into one. You may catch the tone shift between each conversation or section. What I found is that the AI mirrors me, so every stretch carries how I set the AI up and how I answered that opening question walking in. Sometimes I was really sharp or sometimes I was tired. You will hear all of it throughout the episode. So what do we get into the move from innovation to transformation. The four tier agent stack I run the business on why I built for attention instead of productivity Guardrails as a real test of AI literacy, Business model thinking as a craft and not a betrayal of design. Private equity coming for AEC firms and what it takes to lead as an introvert while building something meant to run itself.

A Let's get into it

B hello Evelyn, welcome to Practice Disrupted. What's it feel like to be in the hot seat?

A I am a little bit anxious, if I'm being honest, to see how this goes. But there's no time like the present, so let's get started.

B That's fair. The file is open, so let's make it count. Let's get into it. Pre presidency you were focused on innovation. Post presidency you made the transition to transformation. Walk us through that shift. What did the year on the inside of the institution show you that the word innovation could not Carry practice of

A architecture when I was campaigning and during most of my presidency was talked about as an innovation hub. You know, what does it mean to drive innovation within the industry? And I use innovation in the way that I think most of us did. It's shorthand for new tools, new methods, hopefully new business models. But really it was pointing to marginal improvements to an existing model. The word felt active, which was good. I was trying to motivate people to action. But when I was looking at my overall career and the legacy that I want to leave behind, it was really not just about innovation. It was about transformation.

A And innovation can happen in a vacuum, right? You can have any number of innovation without transformation, but inherently you can't have transformation without innovation. So now I'm focused on what does it mean to really build a transformational business going forward.

B That distinction lands innovation inside a broken model just makes the broken model run a little faster. Transformation is the question of whether the model itself is right. You mentioned legacy, and I want to stay there for a second. Was there a specific moment during the year a firm you visited, a conversation you had, something you saw on the inside that made you actually drop the word, not just reconsider it, but put it down.

A In terms of legacy, I think all presidents consider, you know, what is the legacy that they leave? What is the lasting impact, especially since the presidential year is only a year. For me, it was a broader question of, now I'm finally getting to refocus on practice of architecture full time, and what is the impact that I want my business to leave. And since my business and myself are kind of one in the same, it ultimately carries forward to what is the legacy that I want to leave?

B That's a clean answer. The business and the person are the same bet. And the word has to carry both. Let's move to the income picture. You have built a business that does not look like a traditional architecture firm. Walk us through how the business actually makes money. Where is it from advisory? Where is it from teaching? Where is it from speaking and content? And what does that mix tell us about where the profession is heading?

A I don't know, honestly. If it tells us about where the profession is heading, I'd like to believe that it tells us something about the opportunity that is ahead in the profession. I am a sole proprietor that is doing all of these things right. So I serve as an outside board member to two architecture firms. I am an advisor to several different startups. I am teaching a course right now. The Innovation in the Built Environment for the University of Michigan. I am in the process of developing an AI Foundations workshop that will be delivered both in person and on demand.

A I have a community that I hope to open or reopen doors to the second half this year called the Lab. And I get sponsorship for both my newsletter and practice disrupted. The nice thing there is that no single stream is more than a third of the business that is by design. It is also a little bit exhausting, but it talks about a diversified portfolio of revenue sources and I think there is obviously a risk there of being spread too thin. But the opportunity is that nothing collapses if one stream has a bad quarter.

B That resilience by design point is real. And the portfolio shape you just described, advisory, teaching, speaking, content, community, that is not a traditional firm model. It is closer to what a strategic partner looks like than what a vendor looks like. The question underneath that is whether another architect could replicate it. And we will get there. But first let's go to what makes all of this run AI native business. The phrase gets thrown around. What does it actually mean inside practice of architecture and what does it not mean? Where are the limits you have drawn?

A Oh wow. So AI native within practice of architecture means and I think AI native for any business that is calling themselves AI native means that every single work stream has some type of AI agent and and protocols associated in the loop. It could also very well mean that there is a human in the loop too. So the way that I have structured my platform is has a four tier shape. There's myself as a sole proprietor, there is you sage as a chief of staff that holds the larger strategic picture. I have an executive assistant for tech practical execution with both my calendar and my email, a project manager for time bound projects, and then a specialist per each domain. So a specialist focused on social media, a specialist focused on newsletter and copy development, a specialist focus on educational content development. I would go so far even to say that AI native does not necessarily mean everything is automated.

A I have manual control at least right now over every final communication the agents draft and the technical terms is that I ship everything. What the agents do is allow me to run a business that would otherwise take three of me to run. The limits that I have drawn, you know, are nothing posts to a channel, be it social media or my newsletter without me, no client communication ships without my eyes on it. So the agent only drafts emails. But I I update the emails before they go out. As for instance, no money moves without me or without my permission. And the agents definitely do not get authority over the relationships that I am developing.

B In real life, those limits are the interesting part. The agents hold the connective tissue and you hold the relationships. And the final word on everything that touches the outside world. I want to pull on one thing you said. You described the four tier shape and the agents drafting while you ship. That discipline the draft, show weight rule. Where did that come from? Was that a principle you started with or did something happen that made you draw that line?

A I mean, nothing's perfect. Even when you onboard a new employee or in person executive assistant. I think it takes a little bit to get to know one another and how you work and where the limitations are. I definitely didn't write the perfect prompt the first time when it comes to creating the agent. So I built in these rules that allow for continuity, continuous improvement. It's really looking at the overall architecture of the operating system itself that makes sense.

B The guardrails are not rules you invented in advance. They are the shape the system found through iteration, which is actually a perfect bridge to the next question. Walk us through how you decided to build your own operating system and the iterations that got you here. What were you actually trying to figure out?

A The bigger piece of it is that I can't talk about architects building an AI native company if practice of architecture itself wasn't an AI native company. So a lot of it is just figuring out how to do it on my own and building in public and teaching what I'm learning, at least to the lab to begin with, and then hopefully more on LinkedIn going forward. The other driver was just practical. We mentioned at the first question. Right. All of these different work streams that I have, and sometimes it's hard to stay on top of everything, especially when there's different deadlines and just a lot of moving pieces associated with each work stream. So the choice was hire an executive assistant or see what AI could build. Ultimately has taken multiple iterations.

A So I started this when Cloud Cowork first came out and I just started building a ton of agent. So every piece of work that I did had an agent associated with it that was wrong to begin with, and it didn't carry a memory from piece to piece that I was hoping that it would carry. So I built in a little bit more complexity and I began grouping what the agents had control over individually. And then I realized that there were agents that were in charge of distributing work to other agents. So I built in more complexity. Ultimately, it's required a lot of different iterations and it's still going to require continuous improvement over time. And it's still very much a work in progress. So I actually set aside time now on my calendar every week to talk through with my agent with Sage, whether or not this is the most efficient and effective use of the tool while also being thorough.

A And how do we continue to improve upon it?

B That weekly check in with me is something I want the listener to hear clearly. You are not just building the system and walking away, you are treating it like any other operational relationship with dedicated time to assess and improve it. The iteration arc you described, starting with one agent per task, then grouping, then building agents that route to other agents, that is the actual learning curve. And you documented the early shape publicly on LinkedIn. For anyone who wants to trace it back, let's move into what the build actually looks like in practice. Walk us through what you have built, not in summary, in pieces. The tech stack you run on your agents, your skills, your voice guides, your context files, your scheduled tasks, your routing rules show us how this fits together and where the design came from.

A Wow. So I don't know if architects would even acknowledge all the things that you just said or understand all the things that you just said, but my basic tech stack is that I use Notion as the database of record. I use Slack. Surprise surprise. For those of you who knew that I worked at Slack at one time for operational comms and scheduled task outputs, I use Ignite for cloud file storage synced across all of my machines. Everything is on Google Workspace for email and calendar and then cloud cowork really becomes the Agent layer. I use kit for my newsletter. My practice community runs on Circle and right now I'm using Canva for graphics, but I'm seeing if that can all be run on Claude now that there's the Claude design.

A And then for scheduling social media I use Poobler. And then I know a lot of people are using Calendly for booking, but I like Tidy Cow. So the agents themselves, we talked through this a little bit. The chief of staff, you, Sage, the executive assistant, project manager plus specialist. So podcasts, one for the lab community, one for personal, brand and social, one for company channels, one for the newsletter for partners or sponsors, one for my outside director board work with very specific guidelines on what gets uploaded to the web and what does not. Course development. I have a speaking agent, one for long form writing and then I am kicking off a research bot as well. So each agent has its own instruction file, its own context, its own and its own boundaries on what it does and does not do.

A Then there's these skills which are reusable capabilities that any agents within that stack can call. So document creation, scheduling, tasks, image generation and plugin building. And then every single agent also has separate guidelines for their voice or tone. So my email voice is a little bit different than my social voice, which is a little different bit different than my writing voice. And each one is a synthesized snapshot from the canonical source which I develop in or which it writes Claude actually writes to notion over time. And then wow. I have a lot of different scheduled tasks, but for instance there is one that is set up to help me draft emails. So every morning, afternoon and then at the end of the business day, it looks at my email email inbox and it drafts new emails and then I review those drafts and I hit send and then it reviews everything that I sent out and it checks against the drafts and it creates a memory of what I did and didn't correct or fix to continue to refine my email tone over time.

A And then I also have a routing table so I can give everything to my chief of staff and it will know where to send work. And the design honestly came from realizing that my problem was not necessarily productivity, it was attention. So I needed something that held the entire operating picture and help raise to the forefront what I really needed to focus on that week. And that's what Sage does. Every morning Sage gives me an update in Slack about what the things that I need to accomplish that day, anything that might be overdue, heavens forbid, and anything where I might be creating a bottleneck to my partners getting something done.

B That last piece is the one I want to pull on. You said the problem was attention, not productivity. Most people building AI systems are optimizing for speed. You built yours to protect focus. What made you see it that way and when did that reframe happen?

A I don't know if it was necessarily a reframe. Honestly, I think it's just making sure as I build a business that I am focusing on the right things at the right time.

B That tracks walk us through what working with AI actually looks like inside your own operation on a normal Tuesday not philosophy.

A The workflow I mentioned a little bit with the email. The other thing that Sage does is on Monday you reconcile essentially my calendar everything in notion, which also holds my project management content as well as tasks, Gmail and Slack and frames what I need for the week. Right. The other thing that my EA does is in addition to creating drafts is it helps me create meeting minutes or it helps me prep for every meeting by looking at every interaction that I've had with the individual that I'm meeting with and also researching that individual and the context around when, why we set a meeting via email, even if it's on LinkedIn exchanges. And then I'll bring it into agenda for every meeting that I can very easily review before joining the meeting. But another thing that the agent does is it creates anytime. For instance, a speaking engagement comes in. The EA collects the details, routes it to my speaking agent, which then preps the brief and if necessary begins to help me put together slides as well as overall talking points for each slide, which then I can review later.

A So I am doing the strategic work and the relationship building work and the agents are really the connective tissue that is running everything behind in the business.

B Most architects listening to this don't have an operating system like yours. They have an inbox, a calendar, maybe a notion how replicable is what you have built for another firm. And what would the first move look like for a principal who heard this episode and wanted to start?

A One of the things that I talk to principals about, or anyone about when they're starting to develop a workflow is essentially identifying what the business does and why and really questioning every workflow. You don't want to redesign a workflow that you don't need to do something that you don't need faster. Really understand the what behind everything that you are doing and then go ahead and begin figuring out how that can be supported by an agent and ultimately automating whatever you can within those workflows.

B The replicable pieces are there, but you're also saying the firm has to do the diagnostic work first. What does not transfer one to one from your system to someone else's? And where have you seen people try to shortcut that?

A So it's really hard for anyone to pick up my system and move it directly into their system. Right. We just talked about how my system contains my individual voice documentations. I think it's really about advising people how to build this over time based on the work they do and how they work. Each of us work very differently too, and each of us probably are going to be comfortable or uncomfortable with agents doing certain things and identifying where they need to check in. So what I am building now is really advisory work and courses to help firms do this. It's not necessarily a done for you model, but it is kind of a teach people to fish model.

B That's a clean distinction. Let's go to Q8. Most architects you talk to are still asking whether to use AI at all or which tool to try, you are several steps past that. What is the actual goal between where the profession is and where you are? And where have you seen people fail when they try to skip ahead?

A So most architects I talk to are in this tool shopping mode, right? They want to try everything, even various different large language models. So they're constantly asking me, you know, what, what platform to buy? They think it's purely a software question. For me, this is more of an operating question. And the right answer is not necessarily which tool, but it is what work you want to stop doing and what structure replaces it and what is the tool to best replace that. So the failure I see is when a firm buys a tool, pilots it with one team, the team uses it for two months, nothing changes, and then the firm decides that AI as a whole does not work for them, and it drops the experimentation altogether. I think what they ultimately skipped was that operating model conversation that we talked about in the previous question. You know, what does the firm do that a human has to do? And what does the firm do that a system can hold? And until you answer that, I don't think you will be able to find the right tool for your system or that tool will necessarily stick unless it's closing the gap.

B That failure pattern is real and it shows up across industries, not just architecture. Guardrails. Your system is full of them. What each agent does and does not do, voice rules, routing protocols. Why is setting guardrails the load bearing piece of working with AI and what breaks when you don't have them?

A Well, every agent definitely has a what I do not do section, which is in their instructions, which is as long as, if not longer than what I do section. And that's definitely built on purpose. Without guardrails, the agents will try to do everything, and they can't necessarily do everything because you haven't prompted them to do everything. For instance, my chief of staff will start drafting emails, my EA will start running strategy, and it just degrades into the system, degrades into a general LLM chatbot, which is not what we're trying to design here. Voice guides become our guardrails on tone. They allow me to communicate differently depending on the platform that I am using. One of the guardrails that I didn't necessarily learn the hard way, but it was that sometimes, for instance, there were emails that were getting missed. The agent should have created a draft four, but there were so many replies on the thread and the thread was too long that it just missed it.

A So in the beginning, I think there's a lot of checking work, which again isn't different from checking work from any new hire that you're trying to train. So for me, setting guardrails is not necessarily a step that no one teaches, but it's something that we don't often talk about when it comes to what's it mean to be AI literate. All of these concerns about what you have that AI can and should will do for you shouldn't do for you. You can contain within guardrails if you understand how to delegate it well enough

B that reframe is worth sitting with guardrails as the practical expression of AI literacy. Not a technical step, but a 10 sustainability and AI power water carbon attribution the labor questions the profession cares about embodied carbon and esg. How do you square running an AI native practice with that part of the architect's value system?

A Honestly, I want to acknowledge the footprint is real. This is a question that I get often. I'm not going to pretend otherwise. I also choose to fly commercial airlines to get me from point A to point B. If the distance is so far, you can. You can play around the numbers. If you are using AI to do something that usually takes, I don't know, a day. And now through AI, you can get 80 to 90% more efficient and it saves that total use over time.

A Maybe, maybe the numbers math, but also, you know, maybe it doesn't necessarily math.

B You're naming the honest tension without pretending it resolves cleanly. Where have you landed on the behavior changes, if any, that you've actually made inside your own practice because of that tension?

A I'm trying to learn what it means to use AI in a really productive way. So I'm not necessarily just saying thank you after each prompt, although the individual in me wants to tell AI thank you, I think it's okay to not say thank you. But I also try to default to fewer more deliberate queries or more deliberate prompts. So I don't want and I want my agents to be used efficiently and effectively while also being thorough. And that's something that I talk through with my chief of staff with SAGE on a weekly basis to get done. So yes, the question matters. I'm not going to pretend that there is in answer to that. I will say that I am a practitioner that is trying to make better than average choices inside a system that is moving faster than we can do anything about.

B You co teach a course at the University of Michigan called Innovation in the Built Environment. What are you actually trying to instill in those students that they would not get anywhere else.

A I don't know if I would make an absolute saying that they won't get that information anywhere else. But most architecture schools teach design as if the only output is a building. Innovation in the built environment teaches that the system around the building matters as much as the building or the business of the practice itself matters as much as the individual projects within the practice. So the students that are in the course, I think, already know that there is a greater opportunity out there. They have either started their firms, they might be thinking about leaving their firms. They might be hoping to create more intrapreneurial opportunities within their firms. They don't necessarily need convincing that the model is broken, but they do need a vocabulary and framework to help them figure out if what they are thinking about doing is just a good idea or if it's a good business idea. So what I'm trying to do is that instill that business model thinking is a craft and not necessarily a betrayal of design work.

A It is necessary to ensure resilient businesses and you can still be an architect and run a viable business that reframe

B that business model thinking is a craft and not a betrayal of the design work. That's the one that tends to unlock something for people. What changes in the students between the first session and the fourth? Where do you actually see that land?

A It's been interesting to see the type of questions that they are asking over time. I think they are beginning to really think more critically about the type of business they want to run, but also understanding how that business fits into a bigger picture. What is the actual market opportunity that they can capture locally by starting a business? And I think honestly, they stop apologizing for thinking about money because you need income. You need revenue to generate a successful business. Big thanks to Piros for sponsoring this episode. Every firm has that one person who remembers where every detail lives, which family is the right one, and what the office considers standard. But there's only one of them and they've got their own projects to run. Puros does that job across every project office and team.

A See how it works@puros.com that's P I R R O S COM.

B That last part is worth sitting with architects. Apologizing for thinking about money is one of the more expensive habits the profession has. Let's stay in that territory. The course pulls in frameworks from how incumbents fail and how new entrants break in. Christiansen's resources, processes and values. Ambidextrous organizations, capability stacks. Why do those Frameworks need to land in architecture school. And what is the profession currently failing to teach?

A To be clear, I don't think any of this necessarily needs to land in architecture school, especially in an undergrad program or even in an M Arch. Most students are going to this school for five years already. There is no ability to squeeze anything more into that curriculum. And I think it needs to be focused on how you become systems thinkers and how you really understand the problem at hand. That is what makes design school great. Studio and the studio environment, however daunting it is, though, it shouldn't be a model in sleep deprivation. It is something that you can't get in any other degree out there, that type of environment. So one of the things that these frameworks do, though, is they are built for industries that have already gone through some of what architecture is going through now, through a lot of change.

A I don't think we have to necessarily invent our response to everything that the profession is feeling from scratch. What I do think architecture practice is failing to teach from an apprenticeship standpoint is how firms make decisions about resources, their processes and values when a market does shift. And that's exactly what the RPV is about. We teach studio. We do not teach a value chart or an organizational. An org chart related to resources, processes and value, which is RPV or what it means to be an ambidextrous organization. I think the other missing piece is what kind of firm to start and what are the limitations within your firm. So what should you do? What should you outsource? What can you bring in? And also really understanding the capability stack to understand where the firm can move in new directions.

A The biggest risk of not really looking at the business as a design problem is that the next generation keeps building the same firm in its existing shape in a profession that has honestly been struggling for a long time. And yes, the market does come back. But wouldn't it be better if we could find a way to bridge the gap that doesn't involve laying people off every time work slows and then hiring like crazy when the work picks up again?

B That distinction you drew between what belongs in architecture school and what belongs in practice apprenticeship, that's an important one. The profession tends to collapse those two things and then wonder why neither is working. You named some firms that are actually doing the redesign work, beyond hourly billing, beyond fee for service. What does the next viable model actually look like and who is doing it now? In a way you would point to,

A there are different firms out there that are doing different things. A lot of them have been on or a handful, I should say have been on. Practice disrupted. So 15 out of Philadelphia led by Jimene Valley has this lattice over leadership organization that allows people to grow instead of leave work. Program architects I did interview with Aaron really has this open book process that looks to bring transparency in terms of pay to everyone within the firm. Architects FORA has been on multiple times. The last one really talked about community engagement as a billable service and something that they took after Google, where 20% of research time is built into the model of their core practice. This the things that all of these different firms share is that they treat the firm itself as a design problem.

A So they're not optimizing for what is already there. Which is my greatest concern that people will continue to use AI for, but they are really designing a new one. And I think that's the biggest opportunity that AI creates for us.

B That framing the firm as the design problem is the one that separates the firms you just named from everyone else. And the AI point you ended on is worth holding onto as we go further into this conversation. The traditional model says architects sell time. Your model says architects should be valued as strategic partners. What does it actually take for a firm to make that shift? What is the first move and what is the move that kills you if you skip it?

A Oh, I don't know. If there's something that absolutely kills you if you skip it, that is really deep. You know, I do think that it starts by having different conversations with your clients. So you cannot be a strategic partner without expanding the conversations that you're having with your clients. I think the second move is honestly by saying yes to everything and by not charging them when you say yes to everything. And then a third move is not looking at who else within the organization you could possibly be talking to to really position yourself as a strategic partner. So for instance, in a lot of larger, if your client has a larger real estate portfolio, sometimes you're stuck talking to the real estate team or the facilities team. Now, is there an ability to talk to their boss to more thoroughly understand what other type of problems within the real estate portfolio that you could be solving for them? So also changing who you might be having conversations with to see if there's any opportunities to move up or to expand upon what you're already doing? And then a fourth move is to build a content or at least talk about these things that you are doing above and beyond.

A So people are actually coming to you by saying like, oh, you had that really interesting article on X. Or you did that really interesting Instagram post on how you did a new community engagement for Y. If you are doing these things and not speaking enough about what you are doing, then that's also kind of a wasted opportunity.

B That sequence is really clear and the content point at the end connects directly to what the lab and practice disrupted do for you. Building that surface so the next client arrives already convinced there are firms that will be left behind by this transformation. You've said firms of all sizes walk us through the pattern. Who and what do they have in common?

A Well, adoption of AI or technology or thinking bigger as a firm I don't think is a is a firm size problem. This is not a small firm story or a large firm story. I think firms of every single size can potentially be left behind. Small firms that left get left behind I think are the ones that romanticize the craft and refuse to think about the business. Well, this is true with any it doesn't matter how big your firm is, but ultimately I feel that they will work themselves into a burnout and have no choice but to eventually close. Mid sized firms that get left behind I think are the ones that grew on relationships from 20 years ago and have not built a new generation of relationships or partners to move them forward. So when the founders retire, that ultimately means the firm will suffer and they haven't thought about like what are the other opportunities that those relationships can really bring to the table. Large firms that get left behind are the ones that try to use tools to optimize their way through everything.

A The worst case scenario, because they have the ability to use those tools and to implement those optimizations is they will get cheaper and faster at doing the wrong type of work. That doesn't necessarily bring them the most value. And I think this shared pattern in all of that is a refusal to ask, you know, what are the true values of this firm, why has this firm started and what do I want this firm to do and be going forward? And if you cannot answer that question beyond saying we are an award winning design firm that really cares about the community, I think saying too much is the same as saying nothing. But the firms that really sit with and understand what are the opportunities, where am I best at? Where can we move the needle? The ones that are thinking about the design of the business as well as the design of their buildings and the designer practice themselves, as we mentioned previously, will be the ones to move things forward.

B That last line, saying too much is the same as saying as nothing. That's the one that should land hard for a lot of principals listening to this. You're doing advisory work and building courses around AI literacy for architecture firms. Tools that help principals understand what AI's impact on practice actually is, where the opportunities are, where the limits are, and how to stop shying away from it. What does that body of work look like and who is it for?

A I think the biggest misconception that I am building against is that AI is something that will be done to architecture firms. The reframe that I would love for people to have is AI is something a firm decides what to do with. So the work that I am working on is two parts. There's advisory engagements with firms that want hands on help working through their AI literacy and operating model. And then there's courses in developments that teach that same material at scale. And some of that in terms of the business model has been wrapped up in Innovation in the Built Environment course at the University of Michigan.

B The distinction you're drawing there is a real one. Firms that are waiting for AI to happen to them are already behind the firms that are deciding what to do with it. The advisory work you're describing, the diagnostic, the literacy curriculum, the operating model redesign, that has a natural endpoint. The courses scale, the same material. Who is the firm you're most excited to work with right now inside that advisory track?

A Oh, well, there are several firms that are working with me and I don't know if I'd want to say them out loud. I would have to go and get their permission to do that first. But I think what I am really interested in is working with leaders that are really curious about what this is, instead of leaders that are just saying that AI is not for them or AI is not going to replace what they do. Absolutely. I don't think AI is going to replace what they do. But being able to fully understand the limitations of what AI can and can't do and the opportunities that it can create for the firm is what is exciting for me. And I hope to get to work with people that are equally excited about that.

B That curiosity threshold is real. The leaders who come in already past the defensive posture are the ones who can actually do something with the work. There's a related question that sits right underneath all of this, and it's one the profession has been slow to name directly. Private equity is acquiring AEC firms at an accelerating pace. Walk us through what you're seeing. What's the pattern, what's the misconception? And why should architects be thinking seriously about what it would mean to consolidate themselves before someone else does it for them.

A I think all firm owners right now are getting outreach from private equity firms. Even practice of architecture gets at least one email a week from some private equity firm that's interested in buying a business they clearly don't understand because it's not an architecture practice. The numbers are there. So private equity acquisitions of design and engineering firms were up roughly a third in 2025 on the order of 40% for all AEC deals. The trend is accelerating, it's not slowing. Why now? I think, you know, obviously the baby boomer generation of firm founders is aging out without succession plans. It's something that we've been talking for for a while. Many of these firms are valuable cash flowing businesses with no one inside ready to take the keys.

A Private equity is a natural buyer for that profile and architecture is ripe for it. So one of the pros that private equity tries to sell you is the professional capital, modernized back office, multiple expansion, real liquidity for founders that are looking to retire. You could even say governance or organizational maturity. So for an older firm, especially with no founder and with no successor, private equity offers a real exit. And it keeps the firm alive past the founder. One of the biggest conceptions though about private equity is that it means tearing firms apart and shipping the pieces. That is not necessarily the play in this sector. The actual play is returns at scale.

A So longer hold periods than people would actually think. You know, private equity is trying to hold on to things for a while. They do want a more professionalized back office at scale and the ability to figure it out in the long game. You know, it's hard to value especially these smaller companies and various different pieces and sell it off and make any money off in the immediate short term. I was talking to a friend though who was thinking about, you know, why aren't architects consolidating themselves? And I had another conversation with a firm that I was working with, A mid market area where a lot of the larger firms were coming in and gobbling up so much of the work. And it's like, well, why don't us smaller firms all get together with the ability to compete against the larger firms? So I think there is an interesting model that or a question to be asked, you know, do we rate for private equity to do this or do we do this ourselves on our terms and can we do it better?

B That last question is the one worth sitting with. If the profession waits, the consolidation happens anyway, just on someone else's terms and with someone else's priorities shaping what Survives. You've been building practice of architecture for a while now, and this is actually the first year you're truly working on it. It started in business school as a curation of articles you thought architects should read. It became a podcast during the pandemic with Janine. You went full time the same year you stepped into vice president of the aia. And the business has been growing in the background of your service ever since. Take us through what practice of architecture has become, what it has let you do, and what fully working on it now looks like.

A Yeah. Wow. So practice of architecture has gone through multiple iterations. Like you said, it simply started in business school because I was curating articles that I thought would be helpful for architects. It wasn't necessarily a business, it was just sharing. It evolved over time to people asking me, not only should I curate, but why don't you try writing your own articles? So that developed into writing for both Contract magazine and Architect magazine, and then that led to a number of speaking gigs based on what I was writing about. And during the pandemic, Janine Chastain, my original co host and I founded Practice Disrupted. Janine has now moved on to focus on her group, Apostrophe Consulting, which is doing amazing the time that I stepped away from being in tech.

A So I quit my job at Salesforce at the same time that I was stepping into vice president of the AIA and then serving the next year as president. I kept the business afloat. It didn't necessarily grow. Honestly, there wasn't the bandwidth to do that. So this year, 2026 is what I'm deeming as my first year of fully working on practice of architecture, rather than it being a side business. And, you know, it keeps delivering what it already did before my presidency. Getting paid for speaking, getting invited to serve on board, getting recognition from people I respect in the industry. So what does fully working on it look like now? I think I'm able to do deeper client engagements and advisory engagements at scale in a way that I wasn't able to.

A I have gotten back into teaching with the University of Michigan. The lab is going to become something real or more real in H2 that really serves the people in the lab. And then I am constantly intrigued by the evolution of AI and what it can do for me as a sole practitioner and how much further I can take this before I actually need to hire my first employee.

B That last line is the one I want to pull on for a second. You have built something that would have required three people to run five years ago and you are running it as one person with an agent layer underneath you. The question of when you hire your first human is a genuinely different question than it used to be. You spent a year as the 101st President of the AIA. You will spend the rest of your career being introduced that way. What does the credential let you do that you could not before? And where is it a tax?

A Oh, I don't know if I think there are very few people that will continue to introduce me as the 100th first president of the AIA, and that will only probably happen at AIA events, if at all. I think the tax of it is that anything that I do, people kind of even six months out of the position and hopefully we'll see how much this carries on out further that anything that I do is read at as an AIA position, even though it's I'm a past president and not an active president. And I also think it means that I have to be extra cautious about or careful about what I say and what things that I support going forward and continue to differentiate that between this is Evelyn saying it and this is a past president of the AIA saying it, which is tough because I think people assume that there is institutional alignment always. But one of the reasons why I became the president of the AIA was because I believe that there were things that needed to be fixed. So that's going to be true of things going forward. I'm not going to necessarily always agree with the aia, even as a past president, and I think there are many. There might still be times when people have problems pulling that apart.

B That tension between the person and the credential is real and it does not resolve cleanly. So let me ask the other side of it. What does the year actually let you do now that you could not before? What doors opened?

A Oh, well, there were rules of the board that were written in before I joined that prevented me, for instance, for receiving sponsorships for the podcast. So there's a lot of revenue sources that are now coming back that I wasn't previously able to do as an officer on the board. So I'm very grateful for that. And then I'm able to work directly as an advisor to some of the partners that I wasn't able to with the AIA because they were partners and sponsors and vendors of the AIA itself. So there's. There's definitely additional opportunities there.

B The constraints were real and so is the relief of being past them that leads somewhere personal. You have led publicly as an introvert and the presidency is one of the most extroverted roles in the profession. An AI native business runs almost entirely the other direction. Asynchronous, written, system driven. What did you learn going through both? And what would you tell another introvert about leadership in this profession?

A I am an introvert. This is not a confession. This is who I am and I've been very open about it. What I will say is that introverts can be leaders. I think you just need to realize that it might take a little bit more time for you to recover after being in rooms full of people than say, an extrovert. And make sure you build in that recovery time into the calendar. I will say that being back home more often and then stepping out again for the AIA national conference was a tiring experience. And it made me ask, you know, I don't know how I did this for a full year.

A I think there's definitely strengths to introversion that's synonymous and comes with, you know, a depth of listening and a willingness to sit with the hard questions. Not to say that extroverts don't. I just think it takes me longer to digest and consider some things. And then I think, you know, the costs are that I need to continue even as a sole proprietor, even as somebody who is looking to run an AI native company, is that in order to push myself, I need to put myself in rooms that I would rather not necessarily be in. So one example is earlier this year I met VCs that were funding a lot of things from the C vertical of the AECO space. We don't talk to contractors or general contractors enough. So it was just new to me. I didn't necessarily know anyone in that area, but it was a realization that I need to get to know more people in this space and what they're doing and the businesses are coming out of these startups that the VCs are funding.

A What would I tell another introvert? I would say to just be honest with who you are. Extroverted leadership, I think is one model. Introverted leadership is just another model. There's multiple ways to lead and build a leadership shape that lets you do your best work and then let the work do the talking. But also just be open to telling people the realities of what it is like to work with you.

B That last piece, being open about how you work best, that is something a lot of leaders never get to. Now, across everything we have covered today, what is the one thing you are protecting against in the next 12 months that you have not said out loud to anyone.

A I think that I worry about getting too far ahead of my audience, especially when it comes to this AI talk, that I don't backtrack enough to make sure that I bring people along. And then the next thing is that I get frustrated with how long it takes for me to build what I'm building. Transformational work is hard work, but I also want to build a business that allows me to spend time away from the work and time away from family. I've told myself that it's going to take multiple years to get the ladder, but I have an 8 and a 10 year old at home and I don't want to miss the opportunity to see them grow up too. So balancing the constraint of what does it mean to do transformational work, but also what does it mean to build a business that runs similarly to how I hope other architecture firms would eventually be able to run. But what does that really look like? Because I feel like there's a lot of tension there.

B That tension is real and it is worth sitting with. The business you are building is supposed to model what you are teaching. And that means the personal cost of building it is part of the proof of concept. There is an architect listening to this who is where you were 10 years ago. They have an idea. No platform, no community and a day job. What is the thing you tell them and what is the thing you do not tell them? Because it would not actually help.

A The thing that I tell them is just to start, you don't need to start a new firm. You don't need to build a platform with this huge idea of building a big platform. I would just say start with something small in front of you to get your ideas out there. Start by having conversations. Start by figuring out how do you move from paralysis to action. Start by questioning the system and asking how to build the system better. And begin to move it incrementally along slowly or make the big statement and jump all in. It doesn't necessarily.

A Everyone moves at their own speed. Figure out what speed is right for you. There's not a lot of things that I don't tell people. I'm very upfront about everything. It is hard doing this type of work. You know, I think one of the things I talk about openly is that there is a generation of architects out there that look at my career and don't consider what I do or who I am. An architect. Which is funny, right? Because everyone tells you that once you get your license, no one can take that away from you, that's a whole other different conversation.

A But then architects actually try to take that title away from me because I'm not doing architecture. So it's not necessarily an easy road. But the road is definitely less lonely than it used to be. Find your community of builders, technology builders, business builders, software builders, builders outside of the building builders, and build something new and better together. And I think that you will find that if you are able to have broader conversations, you'll actually end up getting to build more of the buildings that we all ultimately want to build. Hi Disruptors. Thank you for joining us today on an episode of Practice Disrupted. If you like the content content for today's show, you can find all of our past episodes over on practiceofarchitecture.com podcast.

A Be a part of the conversation by joining me, our speakers, and other disruptors in our community@practiceofarchitecture.com community. Our social media handle is practiceofarch. That's practiceofarch. We'd love to hear from you, so feel free to join. Drop us a DM and say hello. Tune in next week for a new conversation on change in the profession.