#55
How to Adopt AI at Your Arts Organization: A Practical Guide with Ashley Gross
Artificial Intelligence is no longer a futuristic concept, it’s an operational necessity for lean arts organizations. In this episode, Aubrey Bergauer sits down with Ashley Gross, CEO of the AI Workforce Alliance, to demystify how AI can save cultural institutions tens of thousands of dollars and hundreds of staff hours. Whether you’re an AI skeptic or an early adopter, this episode provides a baseline for understanding the current tech landscape and moves quickly into practical application for cultural institutions.
Transcript
[00:00:00] Aubrey Bergauer: Hey everyone, welcome to episode six of this season six of The Offstage Mic. If you have been following along this season, you know the theme for this season is changing the narrative in 2026. What does changing the narrative mean in arts management at our cultural institutions today? This year, we cannot have a conversation or a season about changing the narrative in 2026 without talking about the topic we are covering in this episode.
This topic is a game changer, to say the least. It has absolutely been a game changer in my own business. You'll hear us talk about that in this episode. It has saved me tens of thousands of dollars a year and hours and hours of time.
Have you guessed the topic yet? It's AI, my friend. This topic of AI in the arts is definitely in the category of was not part of the dialog at all. Ten years ago, when I started blogging about the strategies and tools for arts organizations, for arts administrators and arts managers to adopt to help us buck the declining trends in our field.
I mean, AI was a term like it certainly existed then, but the idea that it would be part of our everyday lives and like literal everyday conversation the way it is now, was unfathomable. Then ten years ago, at least to me, it was, I'll speak for myself. It's also in the category of operational game changer. I already said it's been a game changer for me.
And I say that game changer, operational game changer. Not in the sense of reducing the staff support we need, even though that has been true in my own business to be transparent. But I, I personally think for arts organizations we are not going to lose tons of jobs to AI, at least not anytime soon. Why?
Because we are already so lean, so understaffed, so overworked. All the things. And so I personally do not think that AI is going to replace arts administration jobs arts management jobs anytime soon. Like I said, I also think it's not going to replace artistic jobs anytime soon in my opinion.
But and I should say what it absolutely can and will do for the people and organizations who embrace the technology is it will reduce the burden, meaning AI will take off some of the load of low level tasks and in turn free us up. You will hear us talk about this in this episode. This idea of freeing up our teams, freeing up our staffs to do higher level thinking and higher level work. That is the game changer I'm talking about, and we are here to dig in and learn about how to do it today.
Now, if you are new to AI or new to adopting AI, I should say no worries, we are going to start this conversation at a pretty baseline place so that no matter who you are, no matter your experience or level of familiarity with the technology, can get on the same page, have the same context going in, and then we are going to get into some examples and use cases of how we can apply these tools and this technology to our organizations. We're going to talk about how to think about privacy and policy, and set guide rails for use of AI at your organization. And we're going to share how I've used and deployed AI in my own business, which, as I said, has saved me tens of thousands of dollars. Honestly, at this point, probably a few hundred thousand dollars, if I'm being honest, because of the way I've integrated AI and specifically AI agents into my business over the last year or so.
If you don't know what AI agents are, do not worry. We're going to explain it, define it, unpack it, and tell you how you can use those tools to. And by the end of this episode, you two will be totally well versed in all of this language and vernacular as it relates to AI. That's the goal.
And as we get going, I just want to share a little bit about the evolution of AI in the workplace, just to help give us some framing. This take or this framing comes from Greg Qof. He is the CEO of a company called section, and he said this was back in April 2025. So over a year ago at the time this episode drops is when I heard him speak at this AI summit I was at, and he said that when AI first came onto the scene in terms of like the modern way, we talk about it today.
So thinking like 2023, 2024 or so, he said. AI used to be the equivalent of, quote, a summer intern that never stops working. Those are his words. That was his take, he says.
Now AI has moved up the org chart and you're definitely going to hear us in the guest I have today talk about AI and thinking about it as kind of an employee on the org chart. So he says AI has moved up the org chart. It is now more the equivalent of an analyst. So we're talking more critical thinking I think is where he's going with that.
And I just really liked that explanation of the evolution of the technology. He finalizes this take by saying now it's replacing people on the org chart. But again, I don't think that's true for us in arts and culture just yet because we are so understaffed. But all to say, I do really believe this technology is one of the few ways we can gain or save time, money, resources.
We are such a resource intensive industry here in arts and culture. I talked about this a lot in the first episode this season, and also in that episode I said there are very few ways we can save or stretch our resources. There are very few efficiencies to be realized, and I've spoken a lot about that before. But this is the technology that starts to relieve some of that pressure, accumulating some of that pressure we face at our institutions.
Our guest today is known for her ability to use AI specifically to eliminate busy work. So just know today we're not really talking about coding or vibe coding. If you know that term or we're not going to talk about any of that kind of AI stuff, we're talking about the most basic and practical, in my mind, at least, basic and practical use cases of AI for our arts organizations. Busy work is the stuff pretty much every job has.
All of us have busy work as a part of our job, some more than others. That's kind of the nature of the beast, at least the nature of the beast the way it was and has been. An AI is here to help change some of that for us. And you know we don't like it because busy work is not fun.
It's boring. It's repetitive. That's a keyword. Things that are repetitive are a great keyword to think about.
How AI can help us do that. Busy work is not stimulating. It's not allowing us, in other words, to use to show up with our full potential as humans in the workplace. And what this guest does particularly well is not just automating busy work to save time, but and also she is so skillful in how she translates that to the bottom line of generating more revenue, which every business wants and needs, including us.
Of course, in the arts, we are also going to do a few exercises and examples in real time. So if you already have a tool that you use ChatGPT Claude Gemini Copilot pick your favorite. If you already have a tool you use, I totally encourage you to make this podcast episode a bit of a working session for yourself. Try the same prompts as you'll hear us do together.
See what comes up for you. Use this as a time to just experiment. We learn by doing so. Seek out how these tools can help you and what kind of comes up for you as we do this today.
It's kind of fun. It's kind of nerdy that tracks, if you've met me, kind of nerdy and with the main goal, really, of having this episode be so practical and helpful for you. So one more time. Welcome to episode six of season six of The Offstage Mic.
This is the AI episode that can change your arts management life. Let's do it! I'm Aubrey Bergauer and welcome to my podcast. I'm known in the arts world for being customer centric, data obsessed, and for growing revenue.
I've been called the Steve Jobs of classical music and the Sheryl Sandberg of the symphony. I've held offstage roles, managing millions of dollars in revenue at major institutions, been chief executive of an orchestra where we doubled the size of the audience and nearly quadrupled the donor base, and wrote a bestselling book on the business of the arts. And I'm here to help you achieve all these same kinds of successes. In this podcast, we are sorting through data, research and business strategies from inside and outside the arts, applying those findings to our work and bringing in some extra voices along the way, all to build the vibrant future we believe is possible for our institutions and for ourselves as offstage administrators and leaders.
You're listening to the offstage mic. Support for this season of the offstage mic comes from Evolve Arts. One thing I've seen firsthand is that most CRMs make it really hard to run your org like a business. People ask me all the time, Aubrey, is there a CRM you actually recommend?
That's why I want to tell you about Evolve Arts, an exciting new CRM that is revolutionizing the art space and the only one I've seen built specifically to make it easy to execute. Many of the strategies from my book, Evolve Arts, was created by people who were frustrated with the CRM at their own organization, musicians and board members with engineering and tech backgrounds. They created the tool that they needed, so it's built for modern use cases. Performing arts organizations see all the time.
It's extremely powerful, while still easy and intuitive to use. In less than a minute, you can send personalized emails to first time attendees, identify and target patrons in every segment of the long haul model, and share a discount code with lapsed ticket buyers, inviting them to come back. Their team is responsive and supportive. If you are having conversations about CRMs at your organization, please check out Evolve Arts.
That's evolve without the second E. Evolve the search them on Google for free demos or tap the link in the show notes for more information. My guest today is Ashley Gross, CEO of the AI Workforce Alliance. She is an AI strategist who helps mid-market and enterprise leaders turn AI into a real business advantage.
I first heard Ashley speak at an AI summit I attended in April 2025, over a year ago. By the time this episode airs, and I remember being so drawn to how she was able to distill this complicated territory of AI and make it so understandable for me, a common person layperson at the time, I was a real AI beginner who just wanted to get better at using AI and my business because I could see the potential. I remember after that summit racing to LinkedIn to go follow her, and to my surprise, she was already following me and I was like, I knew I liked this person and also thinking like, how can I get her work in front of more arts administrators? So here we are today.
You should totally run and go follow her too, because her content and expertise in this area is so, so strong, so so good. I went on to consume basically as much as for content as I could, and training that she had available at the time, as much as I could get my hands on. And I am not exaggerating when I say it was transformational to my business in terms of how much time and money she has saved me. I have so many automated processes and just smarter systems in place now, over a year later, and it has freed me up to take on new opportunities and to do that higher level work that excites me and fulfills me.
So we're going to do some real time exercises and activities right here. In this episode, I mentioned that at the top, and you'll be able to see what I mean in terms of how her work, her teaching, her training applies to your organization, to that's the goal. So heads up to get ready for that. I'll tell you more when we get there.
But to finish this introduction, Ashley built $115 million revenue pipeline with AI tools. Before AI or artificial intelligence was even a buzzword. We were going to hear more about that story right out of the gate in the interview in just a moment. But that's how she began in this space.
And now she teaches executives all over how to lead, build and think with AI. Ashley has appeared as an expert in outlets including entrepreneur, Forbes, Fast Company, Fox, MSN, and USA Today Network. Ashley also has a new book coming out called the AI workweek, which is available now for preorder. Ashley Gross, welcome to The Offstage Mic.
Thank you so much for being here. Thanks for having me. So I want to start with how you became such an AI expert in the first place. My understanding is that you got a little bit of a bad surprise.
Or maybe it's like a pretty big, bad surprise at work. So can you tell that story? Of course. Yes, it's my inception story, so I wear it very proudly.
Yes, I have been a marketer for over ten years, and back in 2020, I found out I was expecting my son and a little naive. Ashley went into her office and I walked in and I told my manager, hey, I'm expecting. I'm super excited and don't worry, I have a whole maternity plan ready to go print it out. I really thought that like I went above and beyond.
I even laminated it. So I thought this conversation was going to be very much like, congratulations. And like, I don't know, maybe a target gift card or something. And they were really excited for me and they were like, oh my God, congratulations.
That's fantastic. But we don't offer maternity leave here. And in my head, I did not plan at all for ever having that part of the conversation brought into it. And so when they said that, honestly, I kind of just like, took a backseat and was just going through the motions.
And so I didn't really have a choice because I didn't plan to have a choice. And I needed to figure out how to have a newborn and also work a 40 hour workweek. And that's that's really the question that started all of this is just how do I do that? Because I don't have a choice.
And this was this is back before ChatGPT. Yes. Yeah. So okay.
ChatGPT doesn't exist yet and all this is going down. And then you figure it out and the CMO of your company comes to you and asks like, can you now this is great. Can you now train 100 plus people who had no prior, I think, right. No prior AI training or none?
Yes, exactly. What did you like? How did. I'm like, what?
How how did you do that? Because. And I asked that because I think this audience, you know, we're talking about arts and culture, arts managers of all different levels of seniority. And I would say some people are totally on the AI train, some people are not.
Some the big middle is probably super curious, and I feel like that probably matches kind of this group than you were then. Like thrust in front of and said, train them all. Yes. Honestly, I think it was just more so the intention.
I wasn't thinking about what I was doing in terms of like how other people were preceding me because it was the Wild West back then. So there was very much these rumors and these notions going around that if you were using AI, you weren't really working, and anyone that's ever prompted and gotten a good output can tell you, obviously, from experience, like it's work, it's a skill set that you have to learn. And I just think that having a really hard intention that I'm going to be a mom, I'm going to have a career at the same time, and no one's going to put me in one box and make me pick. I just put blinders on to what everybody else was saying and thinking, whether it was positive or negative.
And I just started thinking about what I have to do every single day that's repeatable. And I wrote everything down by hand. And once I had written playbooks, that's when I went back through again and I was like, okay, I saved X amount of time, X amount of hours. Now how do I do it again, more aggressively.
And I was able because of that process. And by keeping blinders on, I was able to actually tie the work and the productivity gains to revenue. And that's when I raised it to the CMO and I said, we have a gap in our pipeline, and I see this gap, and we're having the same call every single week with sales to talk about this big gap in our pipeline. Has anyone ever tried to take our CRM and query it using a couple of different AI tools, and then brainstorm and ID and form a plan, and then come into a stakeholder meeting and discuss that plan and bring analytics into it to see if that's worked before.
And it was such a foreign concept at the time that nobody challenged the way in which they were doing work. So again, just having blinders on and asking questions and not second guessing myself and just saying, like, I know how to do this thing and I know how to show that I can provide value with it. Why don't we try this new idea and nothing else is net new right now, so why don't we do it? That was really what what got the ball rolling and made people trust me?
Because it's one thing to pitch an idea. It's a whole entire other notion to pitch an idea and offer to teach everybody else how to adapt to it. And then fast forward three months and you surpassed your revenue goal by how many millions of dollars? 25 million?
Yes, I have my bonus deck and I will never get rid of it. It's actually like framed on my wall because they say a little line what you did to get that bonus tech? What did they say? I overachieved my KPI targets by 98%, by tying what I did to revenue and showing that pipeline.
And it's crazy because it was a very simple gap that needed to be filled. And it really wasn't about the technology at all. It was just using it to foster more productive conversations that led to solving those problems a lot faster and just experimentation in general. That is exactly why you're here today.
Oh, I love it so much. Okay, so I want to start, I think with kind of the basics. Like I said, I think we have probably a wide range of military among our listeners today in terms of their familiarity, familiarity with AI, that is. And so can you just talk about each of the AI models out there?
We've got Claude Anthropic ChatGPT by OpenAI, Copilot by Microsoft, Gemini by Google. Can you first of all, am I forgetting any or and then can you just talk through each one kind of briefly just to make sure we're like covering the major models and maybe what are their strengths or what are they known for? Just to kind of lay that groundwork for everybody. Of course.
And I would say so between 2020 and now, I have vetted over 500 AI tools. So I know that there's a million of them out there. But when you talk about the, the, the top players, I always go to the ones that you mentioned. I go to OpenAI, which is still the everything hub.
So if you're if you're nostalgic and you need your tools in a place that you can easily reference them and you want your outputs to sound like you like, that's still the tool that I think of because it's it's one of the frontier tools. So we've had the longest amount of time to yell at it and get it to sound exactly like us. Yeah. And then when it comes to research and the ideas of synthesizing external research and adding our own internal spins to it, Claude is still going to be my number one choice there.
And then when it comes to, I mean, and also to like, this is very dependent on what already exists in your in your workspace, because my answer differs depending on which prospect or client I'm talking to, because I'm always a big firm believer in work with what you have. Don't you know, don't try to reinvent the wheel if you're Google, use Gemini I love notebook lm, I know a lot of actual podcast hosts that will use notebook LM and I know it, I'm sorry to say I don't know. Notebook LM what is that? It's Gemini.
Oh it is. I didn't know that at all. Oh my gosh. Okay.
Yeah. So notebook LM is like the the GPT equivalent in in OpenAI, right. Yes. And then if they're a Microsoft shop, I know a lot of arts organizations are then copilot.
You would say. Exactly, exactly. Okay. So we also have a lot of people listening who work in marketing.
This is another reason I appreciate your story just coming up through marketing and using these tools. We also have a lot of fundraisers. We have a lot of chief executives. We kind of have everybody.
But what is the best tool for and maybe the answer is different. But like for drafting marketing copy, for social media posts, for email blasts, like the different kind of marketing, I don't know if I'm want to call it tech stack, but like, what tools do you tend to go to? So it's a great question. I think for me, I am less reliant on the AI tool to sound like me and more reliant on can play ice in the environment where all of my things that are me live.
And so for me, it's notion. I just think from coming out of the enterprise world and starting my own business, I was like, I'm going to have the most consolidated tech stack of tools that only provide value, and I'm going to challenge this idea that I need all these these very big, heavy tools that cost a lot of money. So for me, notion notion has every single one of my brand guidelines, my rules, my, you know, RFPs. And so I really, really, really appreciate using clod only because it's got that MCP connector.
It pulls in everything from notion for me. And then it will write it back so I can grab my, my brand voice, my guidelines. And I don't even like to use the information that it creates to put out externally. I like to use it to brainstorm because I'm really, really bad at creative thinking.
And so for me, it's like all about that first messy draft. And if I have something, then it's a lot faster for me to iterate on than a blank box. I would way rather be an editor than generating the content originally at that point. Yeah, yeah.
Yes. Okay, next. This is like my series of like, what's the best tool for this and that? Okay.
So that's like kind of marketing stuff analysis, analyzing data, whether somebody is pulling things from their CRM or trying to analyze, I don't know, media metrics or just any kind of analysis. What's your go to usually? Okay. So I would say it's a toss up.
I am a Google environment. So I would say it's a toss up between Clod and Google only because I so I think it also depends on the way you're working. So for me, and I promise I'm not trying to be long winded with my answers, but I love that I go to YouTube. I'm a video person, and so when I'm trying to perform research and really learn, I need to go there first, watch a couple of videos, then I will go do research in Google or and Claude.
And I like Google's ability to take a YouTube video and then break it down step by step for me and directions. And then it will cite the parts of the YouTube video that reference that specific step by step direction. So then I'm watching it once to retain it. I'm getting a step by step direction, and then I'm adding it into my notebook so that I never lose contact.
So it's like having my own internal YouTube history, but with step by step directions and the context in which I was actually looking for in the first place. That for me is my my chef's kiss of selections. Oh, I love this. That is.
I think there's another reason why I like you. That's how I tend to learn and consume as well, where I'm like watching and doing and kind of like, yeah, and that's how you teach too. So and I'm getting ahead of myself. We're going to get into some real time examples here.
But but I think that's how you teach for what that's worth like very step by step. And that's why I was like, I can do this. I can build an agent, I can do it. So for what that's worth, thank you.
On that topic of agents. Yes. What is the best tool for a genetic behavior. And also explain that because I probably not everybody listening knows generative AI versus agenda AI.
So define the two. And then and then what's the best. So generative AI is a sub type of the technology where either verbally or text based you're giving it directions. And in return it is going to act on the command that you gave it.
And with an agent it's got more autonomy. So if you ask it to do something one time and you give it authorization and permission, it can go and execute on that task. But then it also retains memory and it can then go act on it again based on historical context. So I like to describe it as when you're thinking about agents like is this actually an agent?
Because the way that everything is marketed these days, it looks like everything's an agent when it's really not, which adds to this other layer of confusion. It's not. Yes. Thank you for naming that, because I agree it gets quite confusing.
Yeah. If you're ever like, is this an agent? Just think if you stop talking to it today, would it be able to act on your behalf? And would you have to babysit it?
Would you have to make sure it's not sleepwalking away from you without your your knowledge and your understanding? And in terms of my favorite agent tools? Okay. So I would say I really enjoy Blue Prism because you can build automations, you can build agents, and you can also monitor them in the same place.
Love a consolidated tech stack? I'm telling you, if I can get away with it, I will. And it's enterprise grade security. So you're getting everything that you would usually get in your bigger software ecosystems.
But you can do all those things in one place and it makes sense. I really appreciate what you're saying of like, we do not need to over complicate the tech stack. It doesn't have to cost a ton of money. Like this is what every arts organization wants to hear.
So thank you for being the person, the expert saying like, it doesn't have to. It doesn't have to be no out of reach. Okay, so maybe I'm halfway answering my next question, which is should organizations arts organizations that are newer to adopting AI, should they pick one tool and just try to get good at using that one platform, or should they use the best tool for the job? Kind of like what we were talking about these last couple of questions and be multi-hyphenate, like, what if they're just starting out?
What's your recommendation? My recommendation is to find the leak in their processes. There's something that exists in every single process that is either making you miserable and you have to do it every day regardless, or it is truly taking up so much time that you have to take your eye off of something else that could be spent more strategically, whether that's saving you time to produce more money, whether that's saving you time to just produce more experience, where you can sit in on other calls, cross train yourself, bring yourself up to speed with other initiatives going on in the organization. I would say find the leak that is painful enough that makes you want to learn something new, because we don't like learning things that are new and we don't like change.
So you have to find your internal motivation. But then additionally, and my answer is changed over the years, but I would say in today's age, every single piece of our tech stack has an AI component to it. So I would actually argue that you shouldn't go get one AI tool to optimize. You should look at what you have right now and challenge the notion of, are these tools actually serving me still to this day?
And if they have an AI functionality because they probably do. Is it a good value? Is it a good ROI? Can you measure it or is there nothing there?
And then you can go and see what else is out there and bring it in. But work with what you have first and challenge what capabilities exist right now. Thank you for that advice. Find the pain point.
Yeah that hits, that tracks. Okay I want to talk now about how organizations should think about AI policy internally. So how should organizations set the guide rails that allow for experimentation, allow for this curiosity, like we were saying, like the pain point is so great, so how do I try to solve for that? How do they allow for exploration and adoption of these tools and platforms while still being responsible?
This is the question I guess every organization is probably asking, but how should they begin to think about that? Okay, so my answer is twofold because I have lots of hot takes on this, but one having awareness of where your organization is at today. So if you are in a technology or a legal or regulatory role at the company right now, and you are in charge of creating some governance and structure, start thinking about like, where's your organization at right now? Do you have AI tools that are just going to the budget owners and the budget owners of these different departments tomorrow can turn around and buy it because they have the credit cards?
Or are the AI tools owned by it? By privacy? By is there a process? Right.
So with that in mind and just being realistic about where your organization is, I would say one, write down a list of what you can control. And so what you can control is number of licenses that you give, what vendor you go with, whether or not you put an RFP together and form a 30 day proof of concept, how rigorous you vet the vendor, and if you talk to their product team to look at the roadmap, these are things that are all in your control. What is not in your control is how your employees are going to experiment with the tools, what they are or not going to do, and perhaps whether or not you can actually measure usage other than logins and, you know, minutes of session. Because most of these tools on an enterprise level, that's all they can give you is number of logins and the amount of time that each of the employees is spending, which does not give you productivity numbers, that doesn't give you employment numbers.
And so with that in mind, the gray area in between that you really have to hone in on is a red light, green light essentially of what do we what can we allow? What tools have we already allowed our employees to have access to where we had to form a plan of what we're going to do to protect their data, and then what are our non-negotiables? Like what are our red lines that we are not willing to risk? So when I think about red lines with a company, because everyone's automatically like, well, my list is 50, I've got 50 red lines like, no, we can't, we can't do it like it's too risky.
What I like to think about is, do you own the data when it's moving and do you own it when it's still? And as long as you own it when it's in transit and when it's in rest, and at any point in time you can tell that vendor take take our name off, we want our data back. Then you're in a pretty good spot. That means that you have at least autonomy enough to dictate what happens to your data.
Okay. What do you mean by in-flight versus still the data? Yeah. So and I love this question.
Thank you. So whenever your data is at risk, like the two points of entry that you have to watch out for is where is your data physically stored. Whenever you're using this AI tool on a daily basis, where is the data from the vendor that you're using, their solution for actually being stored, and then when it's moving. So whenever you're making calls and you're sending signals and you're sending prompts their way, where is it at geographically whenever it's moving.
Because depending on what that looks like, there's completely different roles and procedures that like where is their data center. Is that what you're saying? Yeah okay. Okay.
Because if it's the EU, great. Okay. That's you're probably pretty safe. They've got locked up.
Yeah. But also there there's other, you know, options out there that are not the EU. And on the flip side, if it is the EU that has your data, great. It's really safe.
But if you're trying to do a lot of use cases with it now you're getting GDPR, you're getting the EU, AI act. So it's like a double edged sword of so much safety. But then your use cases are going to be limited. Yeah.
Yes. Got it. Okay. Thank you.
That of course. So I remember this is the first AI training where I heard you speak. We actually used AI to draft a policy. I don't know if you remember, you've done so many of these trainings, but.
It was like a group exercises, virtual training, and everybody was like, put in the prompt or. Yeah, put this prompt in chat, GPT or whatever model anybody was using that day and and ask the AI to give us a draft policy. So can we do this activity. Can we like reproduce this right here, right now that you did at that session?
Absolutely. Okay. We're going to ask it for three options okay. Because we like to pick and choose.
So let me go okay. Wait everybody Ashley sorry. As Ashley is going through this with us. This is like one of these moments where everybody listening.
You should do this to like, especially for anybody in a leadership position at an arts organization who is grappling with this exact question. Question being like, what on earth can our internal AI policy B, or even anybody who doesn't have tons of authority at your organization but wants to help push forward the AI adoption at your organization? Basically everybody listening. Open up your favorite AI app right now, and I just encourage everybody to do this with us.
I'll do it. Ashley is going to give us the prompt okay. So this is going to be very simple, very succinct. So all I'm going to do is I'm going to say as a leader at AI Workforce Alliance Inc.
And then in parentheses actually put my URL craft a vision statement for AI agent adoption that prioritizes efficiency while lining with our core values of education, inclusivity and accessibility. Give me three options. All I switched out. There was the URL of my company, and then the three core values or pillars, depending on how your company phrases it of what your company actually represents.
And here are my three. So I've got bold and mission forward, operational and credibility driven, concise and quotable. And then it even tells me because I've spent a decent amount of time with it. So I would hope it does, but it tells me what it's best for.
So it actually gave me three options. And it said this is best for LinkedIn bios, speaking introductions, pitch decks, whereas this one is better for external facing use and investor decks and website hero copies. But it all says the same exact message. It's.
We envision a future where AI agents don't replace workers. They elevate them. We build solutions that organizations actually adopt, and we're committed to making intelligent automation accessible to every team at every level so that efficiency becomes a right, not a privilege. Okay, shame prompt.
Here's the one I did as a leader at changing the Narrative Craft Division statement for AI adoption that prioritizes efficiency while aligning with our core values of. And. This is where my three are different. Every organization can have different core values.
Of course, I said excellent customer focus and using technology to elevate our work. Anybody who's listened to like the intro of this podcast, you hear me talking about these things. Give me three options. And so mine are the excellence driven approach, the customer centric approach, the future forward approach.
And I just remember in my notes that day I was like, I want a combination of option two and three. I like combination of two and three. This idea of customer centric future forward. Okay.
And then it came up with this like hybrid version saying and changing the narrative. We embrace AI as a tool to enhance, not replace the human connections that drive success in the arts. By leveraging AI to optimize workflows and streamline operations, we create more time for high value engagement with the organizations we serve. Our AI strategy prioritizes efficiency while upholding our commitment to excellence in customer focus, ensuring that technology elevates our impact and expands our ability to empower arts organizations worldwide.
So that's when it came up for me, and that was just a draft. I just thought, I don't know, I just thought it was an exercise. It was so, I don't know, reverie maybe sounds a little dramatic, but I just thought it was so helpful to kind of what we were saying earlier, get the juices flowing, unlock things. And I was like, yeah, okay, is a total draft of a policy.
There's a lot in there that I really, I really liked so well. And also you're, you're doing something where you're taking core values that are actually meaningful to you, that your company stands for, and you're aligning AI capabilities over them, which is like the big picture for me when it comes to how orgs should be looking at AI is like, don't reinvent the wheel. Don't make a separate conversation. Apply it to enhance what you want to enhance and get you to your goal ten times faster.
Let's say you're running your performing arts organization more and more like a business. Insights are clicking. Dots are connecting. You're hungry for more eureka moments.
But what holds a lot of organizations back is just not having time to assemble the data you need. Bolero is a new tool for performing arts organizations that creates comprehensive, beautiful reports that give you answers to many of your complex questions. Questions like what programing actually brought in new audiences? How much of our audience returned this year compared to last?
What events do people drive further in to see the data to answer these questions already lives deep in your CRM. Bolero reveals these answers so you can stop digging through data, sometimes even jerry rigging across spreadsheets trying to extract what you need. I know I've been there and instead allows you to spend your time thinking, planning, and making decisions. If you want clear reports created with your ticket data in a fraction of the time and at a fraction of the cost compared to consultants, head to.
Changing the narrative and click Let's Talk at the top of the page. Bolero will be at both the Opera America and League of American Orchestras conferences this year. If you're attending, go say hello to Bolero. And if you book an introductory call before this season of The Offstage Mic ends, the founder of Bolero will help you identify five additional questions not already a part of their question Bank to custom build for you in your data.
If you decide to work together, go to. Changing to take the first step. Okay, so all of I think all or most of all these previous questions were really about using AI as a thought partner, I would say. And now I want to talk more about agents.
I had you talk about agent behavior and define it. But how do you like automations versus a gentle behavior? Can you talk about the difference between those two things? Okay, so I love baking desserts.
It's like my my sweet spot, no pun intended. It's my my stress release. So how I like to explain that if it's an automation and agent is if I tell you to bake brownies and I give you the brownie mix and you have all the ingredients, if you go to bake that brownie, it's going to come out perfect as long as you follow the directions. And you had all those ingredients and you use the right amounts right.
With an agent, let's just say I gave you the box of brownies, but you were shortening an automation. If you did that for an automation, it would say, well, I don't have two. Sorry, you can't make the brownies. But I would say, okay, I want a cool, totally fine applesauce works instead.
This works instead. I'll figure it out. Don't worry, you'll still get brownies. And it takes that context.
It takes those directions, whether they're verbal, whether they're text based, and then it takes the context, aka the ingredients, what it's supposed to do with those ingredients. And it comes to its own conclusion based on logic and prediction. Thank you. Okay.
With those definitions, what are some common use cases of agents that you see in the different organizations you work with in organization charts? So actually in the org chart. So say more I see this and I do this myself. I approach agents not averse not initially, but I approach agents as being a type of employment where you have your full time employees.
If your part time employees give your contractors, you have your freelancers, and then you have agents that are in positions in your org chart that are doing the admin work. They're writing the purchase orders and printing the tickets. They are making sure that all of the receipts after being uploaded to your app, so that you can count them as business expenses. They are the people in the org chart in the form of agents that are pushing the papers, so that the real humans can focus on what they do best.
Which is human? Yes. Which is human. Okay.
Yes. Okay. I love this in. At first I was like in the org chart.
What does she mean? Oh like give it a, like a job description essentially. Okay. Yes.
Totally. Totally. See what you're saying. This is my favorite use case.
And I have so many I've trained all of my employees using agents because I didn't have the time to to set aside and just, like, have them shadow me, it would have been the most chaotic shadowing ever. So I will always start with a project or a notebook, and then I will add on to that context so that the time they're spending with me is spent asking me questions that they didn't already have context or answers for, or just asking me how I came to that decision. Because again, I don't want to use my agents to train people on how to act like robots. I want the humans to be curious about the context and why I'm doing what I'm doing, and how I'm raising things, so that they can then think like me and make better strategic decisions on my behalf and be an extension of me.
That is way more what I care about. When I think about the value of my my humans in my org chart. Yes. Yeah, we don't need robots.
We need humanity built into this somehow. Yeah. For organizations, arts organizations beginning to think about agents like I, I feel like our default is like, oh, it could be a chat bot. It could be meeting scheduling, it could be inbox management.
That's all fine, I suppose. But like what what are the tips of organizations are like? Yes, I see what you're saying. Free up employees for higher level work.
Bring the humanity into it. How do I begin? Like to figure out what in the org chart needs to happen? Like do you know what I'm saying?
Like how what is the next step? I guess, like how do we think about it? So in terms of going from deploying your chat bots, optimizing your AI tools to agents, I would really think about what in your department or in your business is running really smooth, like you have a written document of how this work gets done, why it gets done, and who does the work right. So now you have an owner, you have a process.
You have written instructions that can be replicated. I would pull an agent in to do that, but I would think of it in the terms of risk versus ROI. And I look at it as like a little quadrant. So I think about it like this, where risk is at the top and ROI is at the bottom.
And if I'm thinking about the different use cases within my organization, I'm putting them on this autonomy spectrum of how autonomous am I willing to get when when I'm looking at the risk versus the ROI. So some some wins that I would deem non-hazardous are always content because you cannot have a business and not put out content. That's just it's not a thing. And a lot of people really hate producing content, and I get it.
But where I think an agent does a wonderful job is just throwing out, you know, tools and naming this process. But you can be tech agnostic and use whatever and swap it out. Let's just say that you have all of your templates in Canva, and you have all of your brand information and notion, and then you've got buffer to actually schedule out all your social media. An agent could take those templates in Canva with one direction or one theme.
It could auto populate in bulk, create 30 days worth of content for you. It can then run it through your notion to make sure that it's mashing your core pillars, your brand themes. It's even running through like your do not say this word. Do say this instead.
And then it's scheduling it on buffer, okay. And since an agent, it's not just going to stop once the job is done, it's then going to take that data. It's going to take it back to notion and tell you what performed well, what didn't, what were the questions, what were the themes. And now you have this whole entire reference.
And this is exactly what I do. By the way, I have over 365 days worth of questions that come from my LinkedIn to my YouTube. So when I'm trying to create. Net new content, I don't have to ever start from an idea and go research it.
I've got questions that real people are asking me in real time, that I can just turn the camera on and start answering. Where in this process can a human come in? Because I'm thinking of every arts organization. Of course, we're all creating content.
So I my sense hearing this answer is, on one hand I'm like, that is so frickin good, how do we make that happen? And then on the other hand, it's like, whoa whoa whoa whoa whoa. How do I have a human oversee some part of this? So where does the human supervisor come in?
When I first set this up, the supervisor actually was getting the email. So I had my marketing team getting the emails of the schedule so they would have the post. They would have the text based copy, and then they would have the day that it's going out, and they would get an email with a 30 day content matrix, just like a table within an email, and they could approve it. They could send notes back, but every single email that they were sending back, they were sending back to the agent to remember his context.
Okay. And so based off of that feedback, then the agent could then go post. And because I too did not want to play around with accidentally, like what would happen if this went out or that one out, what I did was I had it automatically put out my paid ads on Facebook, on Instagram, and on YouTube, and I connected it to my YouTube channel so that if it was underperforming for a certain amount of days with a certain amount of time and like certain keywords, it would pull that ad and it would put a different ad out based on the keywords that were performing well that day. So it was something that still mattered.
It was very manual. I didn't want to do it, but it does impact my business. And so I did that instead. And then I kind of built up that trust and I was like, okay, you're doing good here.
Now you can have Twitter, now you can have Pinterest. Yeah. Like you would do with do went back to the org chart analogy with an employee like, okay, you've shown you you have demonstrated competency in this area. Let's add the next thing.
Yeah I really yeah I'm really loving this org chart analogy okay. Yeah. So I want to do another activity. This is another group participation everybody okay.
So. I started this idea with you in advance. So all right everybody working session. Here we go.
Another opportunity. Everybody get out your Google doc or your notes up on your phone or whatever. Especially if you are hearing these different use cases for AI and seeing the value in what agents could do for your organization. So I want to talk through the example of how to build an AI agent for a general inbox, hello at info at whatever every single arts organization has this inbox for me.
Just everybody knows this was a course I took from you and built my agent, and it saved me like tens of thousands of dollars. And it was just a it's just a game changer. Talk about freeing up people to do higher level work. And I have to say in this category of like building the trust with this quote unquote employee, I had to do that, too.
I have a human overseeing things still before, like drafts before they hit said. Like, I just I'm not quite there on the trust yet. So just so everybody knows like transparency. So okay, also I want everybody to know this is like free consulting we're getting here.
So I'm really grateful. This is the kind of stuff Ashley teaches and charges for. So we're going to link her website in the show notes. But for right now I'm just so serious.
It's really actionable for everybody listening to this episode that you can start doing right now, right here for free. So, okay, Ashley, if you would talk me through the prompts that you give when you are coaching or working with clients on how to build an agent, and I have my notes pulled up for my inbox agent, and I'm going to try in real time to help, like do this exercise with you and everybody else on your Google doc. That's how it started. We talked about like the notes on the Google doc that we then use to inform the platform and technology.
So okay, that's what we're doing right here. So yes ask me some questions. Okay. So the questions that I would immediately start asking are how many different types of emails do you get on a day to day basis.
Less than ten over ten under 20. On a daily basis. Different types under ten like and I was maybe I'm going to much of the weeds too fast. But tell me if I'm like what I'm what's going through my head is as I was doing this exercise originally, I was like, I get requests.
Sometimes we do this interview, we talked to this outlet, I get requests for a meeting. Oftentimes I don't want to say yes to. So that's a whole other area. And then actually really impressed with how the agent handles that.
Request for client support. All of my clients in my own programs as a whole. There's a whole body of work that's client support. What else am I speaking?
Inquiries. That's a huge category. I'm sure I'm forgetting some, but these are like my top. I would say you're asking like categories of questions.
That's I'm answering that question. Yes okay. Yes. Perfect.
Okay. So then for most of those calls that you're participating, participating on before you get an email, right, let's just say maybe, I don't know, 30% of the emails that you're getting. You've already had some kind of an interaction with that person, whether it's a phone call or a recorded call that's zoom or a Google Meet. Would you confidently say that at least half of the emails that you're getting, you know a little bit about what is the intention of that email that you're getting, whether that's because you know, the person you've had a phone conversation with them before or you have experience with these types of inquiries?
I would say often if it's the general inbox, it's cold outreach and I don't know the person, but what's going through my head right now is you were talking about what are the use cases earlier where you have like great documentation or a great process. And this is one where it's like, I have FAQ documents for all of these different categories that I had been using to train the assistants, human assistants who worked with me. So yes. Yeah.
Okay. I hope I'm answering the question. You are okay. You're doing great.
Okay. So then you have you have your mappings, right? You said, okay, I get a couple different varieties, but most of the time like these emails are x, y, z. I would then say okay great.
Do you have those labels reflected in your inbox? Do you have folders with the. I had to set it up, but I did what Ashley Gross told me to do. And I have that now.
Yes. Like like actually map it like they were sort of good. I think to begin with I'm a pretty organized person, but I was like, oh, I see how you can operate really like dial this in a bit more. Yes.
Okay. Perfect. So you just gave a potential agent already some context because now it knows how you look and score those emails and actually put them into their own boxes. Perfect.
So then in your Google Drive Microsoft. Yes. Google Drive. Yeah.
Okay. Great. In your Google Drive, do you have historical references for past emails that you have sent, or do you have briefs for how you like correspondences to be sent? What I have and what I used is with yes, I had some previous examples.
Same thing. It's the stuff I would give past humans. And then I have even an FAQ doc. Yes, it's like the factual information.
And then sometimes it's like. You know, make sure you say this with warmth if it's client support or I don't know, I'm like kind of riffing but like those kind of cuz I have as well. Yeah. Perfect.
Amazing. So then if I were to say okay, amazing. Every time somebody sends you an email, there are automatically going to be put into a pathway. You could confidently say, okay, great.
I've already defined the certain pathways based on the varieties of emails that I get. I'm comfortable with the idea that not all emails should go down the same pathway, right? Yes, yes. Okay, so then if we created three paths, let's just cherry pick.
Let's say it is for speaking engagements. Let's say it's cold. You said cold emails okay. And then let's say what's the third one.
What's a really good third one that you run into a lot. Prospects. Just general inquiries on. Yeah.
Yeah. General. Yeah. Either client and potential client inquiries, maybe, or they want help, but I have to.
I have to direct them to what program would be best for them. Yeah. They want a service. Yes.
They want something. Okay, okay, I see this is why you're really helpful. I'm like, what kind of people do I get? Okay.
No, this is the whole entire. If you can explain this and talk it out loud, you can program an agent. That's. But I feel like this was the exercise when I, when I did your program on this.
Yeah. Yes. So we have our three paths we have where we need to direct them based on those three varieties. Right now, when we put them on these paths, what is the first thing that we want them to experience so that their first experience of us is a positive one, but it's also automatically saving us time.
So if I was thinking about speaking engagements for myself, at least whenever I have an email come in for a speaking engagement, I automatically send an email back and say, hey, I would love to consider speaking on your podcast. Here's how I like to operate. I like to have the questions beforehand. Here's my headshot.
Here's my bio. Are there any upcoming initiatives that you have that you'd like some support on, or is this just a fun speaking engagement? Do you have something similar in terms of yes, I have, sorry, I'm like, yes, I do. I have like I call my media kit deck like this PDF deck and it's all the same kind of similar stuff.
Here's the here are the top three talks I do that I'm asked to do a lot. Here's the sliding scale of my speaking fees depending on budget size of the organization here. Yeah. Headshot, bio, all the things social proof quotes from you know, it's like a pitch deck kind of.
But also information for the person who's asking. Yes. Okay. So that's the one I would start with.
So out of these three paths I would pick that one first because we have the most defined path and the ways in which we want our end user to experience us while they're going down that path, which is the most important aspect, because that's all they're going to see. Really good agents should not give away that they're good agents. If you know that someone's using an agent, they're not using it well, they've already failed. And it's not like I totally sorry to interrupt.
I totally agree on the other side. Months now, probably a year later after, like, doing what you told me to do and having I'm like, yeah, this. Like I said, I still have a human approve everything before it goes out, but whenever I hop into that inbox, I'm like, shoot, okay? Like, they actually are, like rocking and rolling on this thing.
Yeah. Yes. Yes. Absolutely.
If you have an experience where the user doesn't even know it's an agent, they just feel like they're taking care of the whole entire time. You one because you cannot you're not an octopus, right? Like you don't have eight legs. Unfortunately, you don't have the time or the bandwidth to be like, everyone's happy, we're great, we're good, we're hydrated.
Like, so that's what you're using AI agents for it. So then when we get to this next part, the questions I would ask you are okay, so let's look at the things that are static versus dynamic. So you mentioned you have got headshots. You mentioned you've got your bio.
You've mentioned you've got testimonials. Okay. Out of those three items that you're giving to email responds whenever they are asking or they initially send you an email. What is the 1 or 2 maybe of those resources that are dynamic availability, calendar availability?
Yep. Okay. So if you didn't already have a cowardly or another calendar app, what I would then say is let's start by defining what is dynamic and needs to constantly evolve and change, and then what can stay static. And I would argue that if you're connected to Google Drive, you can drop 50 headshots in there and you're probably all the all the assets that's static, that's done.
We got it. Here's the here's the link. Yeah. So I would just say your calendar link then and your customer testimonials because you want to eliminate the manual work.
So you don't want to put those back in your drive. So then what I would do from here is I would go for an N because N18 is an automation tool with agent capabilities. So it's very much a guardrail experience of using agents. So I would set up my automation so that every time I get an email, the first action I'm doing is parse JSON.
It's going to go through that whole entire email. And I'm going to add into that parse JSON node that I want it to search for keywords. And the keywords are going to be podcast interest in speaking. Would love to have you as a guest.
Right. And so if it says any of those words automatically, it's going to say, I think that this person wants me to speak on their podcast or they want me to speak at their conference, and it's going to create. If then branch and I would send them to this if then branch and it would check to see if your email automatically labeled it in the folder as speaking engagements. So now I've got a keyword search and consolidation happening.
And then I've got your definition of what speaking engagements are. So if it's labeled then you are then the human in the loop unknowingly because you've validated already the speak engagement. So then continuing on through there, what I would do is we have two checks. The next check would be go pull a headshot, go pull the bio, and then I would use a third party tool for reviews to run reviews and choose the top five most recent reviews.
And then I would also go grab my cowardly booking link because that's the tool I use. I would make four different calls and then I would add a function node, which is just like a code node. But in 19 I really like the way that they have the set up, because you don't have to know how to code at all. Or you could be the most developed coder ever and it's still going to wear perfectly for you.
You can add a function node, and you can combine all of those different assets with all those different calls into one output, and you can then call an agent, choose the large language model that you want. Right. So let's just say for this instance we want Claude, our large language model, a genetic node would then take that information, put it into the email, and it could be CCU on the email and send that email back to the guest. Or it could send that email to a high at partnerships email.
And then from there, you could have your human look at 25 emails at once every day that they have read the drafts, then send, then change. Something said yeah, okay. Yes. Okay.
Two part question what are you keeping an eye on as you look ahead? And how do you stay on top of it all when it is changing so fast? When you said, you know, you've experimented with over 500 tools, I would just like jaw on the floor. But anyways, okay, that's a lot of words, but what are you looking at?
How do you stay on top of it? Okay. Great questions. So I don't try and stay on top of it because it's overwhelming.
So my rules are I have a perplexity space among other scheduled tasks happening in different apps just to compare the outputs and values, because I like to nerd out and tinker, but I have a place with schedule tasks happening, and it is automatically going through all of the AI tools that I'm using on a day to day basis. So it has my tech stack, and it is telling me whether or not the data policy changed or a feature changed, so that if I have to pay attention to anything about that tool, I know 100%. It's either disrupting the way that my clients are going to be able to experience the tool, or it's actually going to disrupt my own personal workflows. And either way, that to me is a signal that I it's not a waste of my time to look at whatever that is.
You have to you have to put a bubble around yourself and decide what you're going to pay to pay attention to and subscribe to because it's too much, it's too much, it's too much. Okay. To me, that goes right back to the literacy piece. If you are literate at some level, then you can make these choices.
If that's working for me, that's not this is better. That's not okay. Okay. Last question.
Well, sort of last question. If you could wave a magic wand, what is your wish for organizations, arts organizations or otherwise who are not using AI? Or maybe they're interested, like we said, but not aren't just like aren't sure where to begin, or even those who are rightfully afraid of the technology. Like what is your wish?
For those that are, let's call them on the fence. What is your wish for those organizations? My wish is that they learn from history and they learn that it's not too late, and that just because they are perhaps hearing the worst use cases and the horror stories of AI, that doesn't mean that that's all it is, right? A couple stories don't make up the whole entire narrative.
And also, we are not even close to seeing the real potential of this technology because we're not evolving the conversation. So it really does take everybody trying these out, because it's not about the tools, it's about the people that are operating and changing their processes, having conversations about how we can work better, that actually work for us. It's a good answer. Okay, last question for real, where can people find you so they can get more of your content training and teaching?
Yes, LinkedIn, I am the most active on LinkedIn. Absolutely. Please come find me. I will answer any question that you have.
All right. It's actually gross. And she has a new book coming out I want to plug for everybody. It is called the AI workweek How to Cut hours, automate the busy work, and focus on what actually matters.
It's basically everything we've been talking about today and I already have my copy preordered. You should get your copy preordered as well. I just think this is such a great roadmap for arts organizations. Go to her website.
It's all there, everything on the book, and I'm just really grateful to have your expertise here today and how you've just so generously showed up in offered so much expertise to this audience here in arts and culture. So on behalf of all of us, I just want to say thank you so much for that work and for being here today. Thank you for having me. That was so nice.
Hey, off stage, as we've said this season of The Offstage Mic, running an arts organization is hard these days, without a doubt. But what if there was a way to make your revenue goals a little easier to attain, balance the budget and end the year in the black, and feel like you have a real plan for financial sustainability? As I've also said many times, hard does not mean impossible. That is why I am teaching a free masterclass where I share my top four strategies to grow your audience, expand your donor base, and ultimately increase revenue.
Things you can do immediately to start seeing results. Things that don't require a huge team, mega resources or a giant budget. I have now seen these strategies work with over 300 individuals and organizations I've taught in my run. It like a business academy across budget sizes, geographies, and artistic disciplines.
Some of who you're hearing about on the podcast this season, and I'm going to show you how to do it to doors to my Run It Like a Business Academy are now open through June 11th, 2026. But your first stop should be to reserve your free spot and my Audience Growth Masterclass, where I share my top strategies as well as information about the program, all for free. Only once a year. Think about the relief and freedom that lies ahead when you start increasing your sales and donations over just the next few months, that can absolutely be your reality.
Reserve your free seat for this Audience Growth masterclass right now by going to Aubreybergauer.com/masterclass That's all for today, folks. Thanks so much for listening.
And if you like what you heard here, hit that button to follow and subscribe to this podcast. And if you've been around for a few episodes now, would you please consider leaving a quick rating or review? I really can't thank you enough for your support in this way. To all of you.
One more time. Thanks again. I'll see you next time on The Offstage Mic. This season of The Offstage Mic is produced by me, Aubrey Bergauer.
Additional production support as well as editing is by Morreale Digital. Our theme music is by Alex Grohl. Additional support this season comes from Sandy Kobashi and Johan Dewi. This is a production of Changing the Narrative.
This podcast is brought to you by Morreale Digital. We all want our arts organizations to reach wider audiences and engage our communities, right? Well, a video podcast is a great way to do that. But as I know too well podcast to take a lot of time and technical skills to do well.
That's why I partnered with Morreale Digital. They handle the technical elements of production, editing and distribution so I can focus on showing up and sharing ideas that matter. To learn how podcasting can help you reach and engage your audience. Visit morrealedigital.ca/podcast.
That's morrealedigital.ca/podcast. Or hit the link in the show notes.