The AI-era means having in-house expertise

Most leaders want to harness AI—but few understand what it really takes to make it work across an organization. In this episode, Micah and Alane break down what they call the Upside-Down Pyramid, a simple framework for transforming your business with AI-driven systems that actually scale.

They walk through the five stages of transformation:

1. Align leadership – get everyone speaking the same language about AI and automation.
2. Build internal champions – develop in-house experts who can move fast.
3. Establish your foundation – why ClickUp, Asana, or Monday are essential before automation can stick.
4. Design and deploy – how to map real processes (not assumptions) before automating.
5. Keep learning – staying ahead in a world where tools change every week.

If you’ve been wondering where to start—or why your early automations aren’t delivering—this episode gives you the roadmap to build systems your team will actually use.

Disclosure: Some of the links above are affiliate links. This means that at no additional cost to you, we may earn a commission if you click through and make a purchase. Thank you for supporting the podcast!

For more information, visit our website at biggestgoal.ai.

Want more valuable content and helpful tips?

Explore our Courses/Programs:

• Complete our self-paced Process Mapping course

Enjoy our Free Tools:

• Free Asana, ClickUp, or Monday.com Selector Tool
• Get 25 Free Custom Automation Ideas for your Business

Connect with Us:

• Join our Community
• Follow us on LinkedIn
• Subscribe to our YouTube channel

Transcript

0:00:00 - (Alane): Welcome to Automate Your Agency. Every week we bring you expert insights, practical tips, and success stories that will help you streamline your business operations and boost your growth. Let's get started on your journey to more efficient and scalable operations. Micah, you created this upside down pyramid that I think just really paints the picture for people on what an AI transformation. But also just what does it really look like to build scalable systems at the core? I really think that's what it is. Because all of the stuff you have in this pyramid, I think relates not just to AI and how to implement it, but just how do you build systems within your company that people can follow?

0:00:48 - (Micah): Yeah. And I would even clarify that, that it's how do you build systems today? Because building systems 12 months ago is different than building systems today with how fast AI is changing things.

0:01:05 - (Alane): Yeah. So I thought we could talk about your upside down pyramid.

0:01:08 - (Micah): All right. Sounds like a medical condition, but I'm game.

0:01:13 - (Alane): All right, so at the very top, and we've got this shared so that if anybody wants to take a look at Micah's, you can look at it on the video or we can also put it in the show notes. So the top of it is the leadership and making sure everybody's on the same page.

0:01:30 - (Micah): Yeah. So we see this all the time, especially with AI. We'll talk to team members who are our leaders, their managers, maybe in their own departments or their own business units. And they're getting instructions from their bosses. Maybe it's even the board members who are going to, hey, do that. Can you, can you AI this? Can you. Can you just, you know, add the AI and these team members and these managers and these leaders are going, what the F, man? Like, what do you mean, AI it? What do you actually want me to do?

0:02:09 - (Micah): And the response is, you know, like, it generates the stuff and it solves the problems and it does. It just does it automatically. And there's just this huge disconnect. And so there's kind of three, three big categories on this with leadership that we see that really needs to be level set. And one is how the hell does AI work at a core level? Not all the details, not all the technical stuff, but what is actually happening so that everybody can speak the same language and know the limitations and know what's realistic and know we haven't invented Skynet yet.

0:02:45 - (Micah): The second one is what are agents really? What are agents? What can they do? What can't they do? What are the three pieces that make them up? Right. Like, what are the Components and how do we start using these things? How do we make better decisions around all of this? And the third one is RAG and vector databases. How do we go beyond the AI's context windows? How do we leverage our current company assets like SOPs and handbooks and pricing sheets and all the other assets that are out there and start removing roadblocks. And so we've been doing workshops with leadership teams on these three specific topics and it is mind blowing because we, you can just watch the light bulbs just start lighting up in everybody's head. And the real key, and this is why it's at the top of the upside down pyramid, is this means we can be comfortable with AI, we don't have to be scared about it, and we can make appropriate decisions and work with our teams on where we can go from here.

0:03:52 - (Alane): Yeah, I mean, a lot of times we'll go in and the leadership team thinks of AI, and I shouldn't just say them, but people think of AI as what is accessible to me right now. So ChatGPT, Claude, you know, something where it's easily accessible through chat. Not necessarily understanding what RAG and agents can do and so really making sure that leadership understands these capabilities and how it can be impactful for their teams. Because we're not talking about one on one manual chatting. Like to me that is still manual work because it's isolated between you and the other person and it's up to you to write that prompt.

0:04:36 - (Alane): But when we talk about agents, that's no longer a one on one relationship. Now we're really talking about automated work throughout your departments and organization.

0:04:45 - (Micah): Yeah. And I think, you know, if we, if we think about this one on one thing, it's creating silos, it's creating the wild west of operations. Well, John over here, he created his own like agent mode in ChatGPT. You know, Barbara over here, she created some cloud skills. This is two different systems, two different providers, two different agreements, two different things that are happening. Nobody else can access them, nobody else knows what they're doing. It's only individual to them. You take that across an organization and you are in deep trouble trying to reel this all in.

0:05:22 - (Micah): And so, you know, that's part of level setting. The leadership as well is understanding, well, are you just letting the wild west happen or are you being constructive about it and saying we need to operationalize this, we need to standardize this, we need to build true systems and I'll.

0:05:42 - (Alane): Make one more point on this and then we can move on is what happens when those Employees leave.

0:05:47 - (Micah): Yes.

0:05:48 - (Alane): And you didn't have an AI policy in place with paid platforms and they leave with your company data. So understanding that those implications from your leadership team so that you know, hey, we got to put an AI policy in place and we need to make sure we are communicating that to our employees and they understand which platforms are paid for by the company that they can access and no other platforms should they be using.

0:06:17 - (Micah): Yes.

0:06:17 - (Alane): Or personal accounts.

0:06:19 - (Micah): Yeah. And we typically can accomplish this in about three hours. So we can cover all of those bases in about three hours. Light bulbs turn on and at a bare minimum, you have your leadership team going, does that mean we can do this? What about this? And a lot of times we hear feedback after we do one of these workshops where they go back into their day to day and all of a sudden everything looks different, everything sounds different. Why the hell are we doing it this way still?

0:06:53 - (Micah): We could be living in the AI era now that we know how all this works. Okay, so moving on to the second one, building champions of in house experts. So, Alane, this one has been, I wouldn't say a point of contention for us, but I think we've had to dig deeper into our own thoughts and our own decision process on how do we want to run as a company. So we do a lot of implementation for our clients and that's pretty much how this business started. We would go in and we'd build systems and we do all these things.

0:07:33 - (Micah): And as AI and AI agents and RAG databases and all this stuff came out, we had to take a step back and look at it and go, is this really still the best solution for our clients? Are we providing the most value for our clients? And I think we ultimately got the answer. That is, there's no. Because we looked at it and Alane, this is, you know, something you and I were chatting about recently is the speed at which we can develop our own tools is lightning fast.

0:08:08 - (Micah): We are flying at light speed. Because you can come up with an idea, you can share it with me. I know how to build agents and RAG databases. I can have a prototype ready within an hour. You can look at it and say, well, what about this? And what about this? I can change it. And that's two people who knock something out and have a working solution within a day.

0:08:30 - (Alane): It doesn't always happen that way.

0:08:31 - (Micah): It doesn't always happen that way. But I'm. But there are. It is still so much faster than if we couldn't do that in house and we had to pay a vendor to do it. Then they have to understand our workflows. They have to understand exactly who we are, what we're looking for, our objectives, our outputs, our decisions, a hundred percent of the nuances. And that takes a lot of time. So with all that said, we started looking at this going how do we empower our clients to be lightning fast?

0:09:07 - (Micah): And the resolution that we came to was we need to train the trainers at our clients. We need to take our knowledge on systems and AI and RAG and be able to train individuals so that our clients can have those champions.

0:09:21 - (Alane): Yeah, this is really a big part and I don't even. This is why I think it's not even just specific to AI. Anytime you're building systems, you need to have a champion there that loves building systems, whether it's AI or building automation within your ClickUp Monday or asana because not every person is that person in your company. Micah, your explanation just now of how sometimes we execute on an agent is I'm not the one building it.

0:09:56 - (Alane): I cannot build agents. I know we need them. I've got the ideas, you've got the skill set, but that isn't me having to do that. And so if it wasn't you, which not always you're, you're not always the person to build them is, you know, we still go through our process with our team where, you know, Sophia and I might say this is the kind of agent we need go through our own project explanation because if it gets too complex, it's not that we still have to communicate that workflow, but the main thing that I want to get across is you don't have to be that person.

0:10:32 - (Alane): You just need someone on your team who is.

0:10:35 - (Micah): And I think this is such a critical aspect of running a business today, tomorrow and probably the next foreseeable future. Right. Is AI is here. AI isn't going away. Your employees are going to use it whether you want them to or not. Students are using it whether teachers want them to or not. Like it's here. We have to adapt our businesses to be able to stay competitive and work in this environment.

0:11:04 - (Micah): And what I think the main point that I want to get across on this one is the only way that you can do that is by having experts in house and you have to.

0:11:14 - (Alane): Train them because the technology moves too fast. You know, we are experts in Monday, ClickUp, Asana, n8n and Make, Zapier. You know, like we stay true to what we are experts in because we can execute so fast. This is our knowledge. We can come in and help build those Champions up so that they have the knowledge.

0:11:37 - (Micah): Yep, yep. Because inside you're going to want to go, well, could we do this? And you're going to want somebody to immediately, like just you and I having this conversation now, Alane, like, this could be the same as us brainstorming on, well, what are our big operational problems right now? Could an agent do that? Could we solve that? Could we get some of this data in a vector database and serve that through an internal chatbot?

0:12:02 - (Micah): You need that discussion internally to happen. Now, we'll get to some of the other more advanced things. But at its. Even if you had somebody that was just the. At the beginner level to say, well, I know how to start N8N I know how to build a basic workflow. I can build an agent and I can connect to a RAG database. And so, you know, we've. We've started training teams on how to do those so that they have the foundations, they are the champions. They can help make leader, help work with the leadership to make decisions on. Is this possible? Yes or no?

0:12:40 - (Micah): Are we heading the right. Are we thinking about it the right way? Is there an easier way to do this? What features can we leverage in the new models? What are the new models capable of? All of those things.

0:12:54 - (Alane): All right. This number three foundation for scale is something that I keep coming back to in all my communications with. Even with you, when I'm like, debriefing from the day, I'm like, at a core, there has to be a foundation work management system to scale your company.

0:13:14 - (Micah): Yes. So. So talk more about that, Alane, because I think. I think that you, you are very passionate about this for good reason. But describe. Describe what your. Where your head's at with all this.

0:13:26 - (Alane): So I know what it was like 15 years ago, Micah, when we were building our company and we had back then probably five or seven employees and just scrambling with where things were at. And the team called me a goldfish because they would ask me to do something. I would literally walk away. And that never entered my mind again because 5 million other things were stirring around and I would lay in bed at night and then start evernoting myself. I don't even know if that app is still around, but I would write messages and notes in Evernote for myself to remember because my.

0:14:08 - (Alane): My brain finally relaxed. We could have never scaled because we were at a point.

0:14:13 - (Micah): I think we just lost our Evernote sponsorship. Alane.

0:14:19 - (Alane): It was great. I loved it. We just put in some other systems. So at one point you said, we need to start Putting in some foundation systems and we're going to do. I mean back then there wasn't a lot of options. There was basecamp that we were using a Basecamp 2 and then Asana came out and so we really started putting in templates for our work. And as we grew to 30 employees and then we did a merger and we were at 125 employees, they didn't have that system in place. And I saw what it looked like for people to start breaking.

0:14:59 - (Alane): We started putting these things in place. People started feeling better and getting their knowing account, having accountability. But the other. So those things are leading to. If you really want automation to work, you need to have a system, a hub for this stuff to go through or else everything's just willy nilly happening. You need a center hub for your work to be going through. If we look at a majority of our agents, ClickUp is working as a part of a lot of them because where do we want the end result of the work to be validated?

0:15:38 - (Alane): Most likely we want that notification happening in ClickUp so our team can go from there and review the information before it goes wherever it needs to go.

0:15:48 - (Micah): Yeah, I use this example all the time. Alane, sorry to cut you off. You're fine. When, when people are asking about this because we can build an agent. You can build an agent. Any, anybody can build an agent. The agent is going to output something. A human, 99.8% of the time is going to need to review that output. That's where we're at. It's probably the agent has probably saved 80 to 90% of the time consuming steps.

0:16:19 - (Micah): But you don't want to take the human element out. If you don't have a tool like Asana, ClickUp or Monday and you need somebody on your team to review the output from this automated workflow that's powered by an agent. Well, do you send them a slack message? Do you send them an email like that's just kicking the can down the road. And again, the chance of them following up with that, the chance of them not goldfishing it, if I put it in your terminology, is very, very low. They are going to forget about it. It's disconnected from work.

0:16:56 - (Micah): But on the flip side, if you have one of those tools in place and you put it in the correct folder, the correct project, the correct task and leave a comment and ention and automatically set a due date and all of that priority and all that important stuff, the chance that they actually execute on this is extremely high. Like 90 to 95 maybe 99% they're going to do it. Especially if they're trained on how to manage their tasks in these systems, which is so important.

0:17:30 - (Micah): All right, now I'm getting passionate about it because, yeah. I'm like, how do you run a business without these things?

0:17:35 - (Alane): I know. So, Micah, this is where I see this starting to break. Not for companies because they're buying in at the leadership. Yes. They're like, yes, we need champions. And then we really start talking about what it looks to scale. And they don't want to change here. They want to. And I get told so many times, I want the systems you have. I want to have my company run like you are talking about. And then I say, okay, this is what needs to happen.

0:18:06 - (Alane): And they don't want to change. Well, that. Slack is how we communicate. Well, then you don't want it to change and you don't want.

0:18:15 - (Micah): You don't get the systems.

0:18:17 - (Alane): So then doesn't sound like what. What you actually want is systems. You just want to continue chatting. And so it takes work. This is the hard part. Changing your habits and your company culture is hard. But if you actually want to put in systems that work, you have to change. Slacking and teaming on projects and client work is not that.

0:18:43 - (Micah): You know, I think, I think asking why. Why are we slacking?

0:18:49 - (Alane): Because people don't know where things are at. They have questions. I mean, I know you're not asking me. You're being rhetorical, but you know, why are they. Because they don't have anywhere else to go.

0:19:00 - (Micah): Yep, yep. But, and, and also like to the cause, I hear the same. The same objections is like, well, the way that we operated, we put everything in Slack. Cool. But at some point you weren't putting everything in Slack, so then you made the decision to put everything in Slack. It is. Is your entire business ending at Slack? Hopefully not. Or maybe it is if you keep this up in a different way. Right. Like, yeah. So, you know, you have to ask yourself, why? Why are we doing it this way?

0:19:31 - (Micah): Is there a better way? And Slack has its purpose. I mean, we've done a couple episodes on this. Slack has its purpose, teams have its purpose. But there's this huge benefit of building that foundation of operating in a central platform. From collaboration to accountability to dashboards to task to just a location where agents can pull information from and put information. Because it's work. These are typically called work management systems. Not so much Project management systems, which I think has.

0:20:06 - (Micah): Is a big misnomer in all of this, because if you think about it from a work management system, all your work, your collaboration, your communication, your tasks, your projects, your goals, your internal growth and key initiatives, all of those things need to be centralized in these systems and your team needs to know how to use them.

0:20:30 - (Alane): All right, Micah, we can break that out into a whole, right? We can break that out into its own episode, which is what, what basically our podcast started from. But we let's move on to 4. Design and deploy. This is another hard one for people.

0:20:48 - (Micah): Well and so this is kind of where vendors like ourselves can come back in because leadership knows what they want, they know what's capable. Now champions are in there, but they typically have their own roles and jobs too. It's very rare you can just hire an automation expert as an individual role. And in fact that's not usually the best case. You've built your foundation for scale, you've got a sonic click upper Monday.

0:21:15 - (Micah): But now how do you build bigger, more complex things? How do you build teams of agents? How do you like. There's always all these questions that are going to come up. We want to achieve X. That's beyond our champions capabilities or new things came out. You're going to need expert guidance, you're going to need somebody who's always doing this. That's where companies like us can come in because we can offer that support and that guidance and that coaching and that consulting to say, cool, your champions are doing X, Y and Z.

0:21:47 - (Micah): Here's a design that might get you closer to the objective, this bigger project. Here's how you could change this or here's how you could optimize and put that all together.

0:21:59 - (Alane): And this the reason that this can be a hard part too, not just in building but, but in mapping. So we're talking about designing and then deploying in this one. When you haven't looked at your process and you're saying, okay, I want to automate this. Now oftentimes a company has several different ways of doing it because it might be an and and but oh well maybe this. But also you might have four people working that do this type of work and they all kind of have their own way of doing it.

0:22:31 - (Alane): Well now you need to figure out one way of doing it, which is the company way. Well, that is very hard for a company to keep moving it along and staying accountable to figuring it out because most people go, oh, this is hard, I'm going to just give up. And having an outside party walking you through and keeping that moving forward in that momentum happening gets you to deployment.

0:22:56 - (Micah): Yes, yes. And I would love to point out here that if. And we're guilty of this too. But if you start with this before your leadership team knows what's going on, before you have internal champions, before you have your foundation and you go, I'm going to start just building this super complex solution, great. You can, but exactly what you're talking about, Alane, how many times have we seen the design phase of stretch into months?

0:23:27 - (Micah): And it's for all of those reasons leadership is confused or have gaps in knowledge. There's nobody internal that can make a decision to save them their lives, because there's no champions that have been trained on this. There's no foundation for where all the work is happening together. And so when you sit down and start diagramming a process or you start mapping out a process, all of those previous decisions that were skipped now are being forced to be made. And nobody's prepared for that. Because in everybody's mind, when we sit down to design, it is, let's draw a few boxes, let's look at this workflow, then we're going to build it. It does not happen like that until you do the work. Like you were talking about a second ago, Alane, ahead of time to make sure we have this foundation to work from.

0:24:18 - (Alane): Yeah, I mean, there's a reason you haven't looked at the process, but now it's because it takes time. And now you need to put the time in to look at the process. All right, are we ready for the fifth one?

0:24:29 - (Micah): Fifth and final is continuous learning. We are in an era where things are changing faster than ever before. I was having a conversation with somebody yesterday, Alane, and it really, it really emphasizes this in our previous businesses. And I mean, this is my third business. We've been doing this for over 20 years in prior businesses and prior ventures. We could do things like test the market, get some validation. Oh, this is going to work.

0:25:00 - (Micah): Cool. Let's invest some time and some money into maximizing and capitalizing on that. And then we could ride it for two to three years. Because while technology changed, it did not change at such a rapid clip like it is today. Today it's, oh, we got some validation. And by the time you start investing some time and money into it, it's already gone. It's something new has changed. Something new has come out, like AI has advanced, new models are coming out, whatever it might be. It is very hard to stay on top of everything that's happening in the world today.

0:25:34 - (Micah): So having a place to go where people like us who are intertwined in all of this. You have to have continuous learning. So it's not let's level set leadership and let's build champions and let's leverage ClickUp Asana or Monday and let's build one workflow and then we never have to worry about it again. No, Running a business in today's AI era is you have to be a tech company. No matter what industry you're in.

0:26:03 - (Micah): You have to stay on top of what is going on and you have to have the agility to adapt, bar none.

0:26:11 - (Alane): Yeah, Micah, I actually just said to somebody when I was in D.C. a couple weeks ago is every company is a tech company and if you don't see yourself like that, then you might as well just close the doors because if you're not using technology, you're not going to survive. It is, it's never been more apparent.

0:26:28 - (Micah): Yeah, I mean, you know, you could, you could think about for any commonalities in business, let's say your business that's getting RFPs, and you need to respond to these RFPs. You can damn well guarantee that your competitors are going to use AI to help read those RFPs and help write the responses to those RFPs. You can pretty much guarantee that it's going to be an AI that's going to read some or all of those responses, summarize and categorize them and grade them, just like with job seekers and resumes and HR teams. Now, all of that is already happening if you are the company who is manually reading RFPs, who is manually writing the responses and manually replying to them.

0:27:15 - (Micah): Your competitors are doing it in days, you're doing it in weeks or months. The decision is already going to be made before you can even get your response over to the company that sent the RFP in the first place.

0:27:27 - (Alane): Yeah, and what that looks like at first is margins are extremely tight that you're not sure how you're going to continue operating at the level that you are. Because margins are tight, because other companies are starting to figure it out and able to still survive by putting in these systems and agents. Like you're saying, Micah. And so for the continuous learning about the technology itself, Micah, you know, we're doing so much now with cohorts and training on agents, but also looking at this from curating the technology because there's no way that a team can keep up with what their specialty is for their company.

0:28:07 - (Alane): You still do what you do because that's what you're good at. And then having. But you need somehow to curate this knowledge and how to stay on top of it, but in a digestible way. So that's something else that Micah, you know, you've really been championing, too, at the company is being this resource and leader for other companies so that you can get this information, digest it, and then educate our clients.

0:28:32 - (Micah): Yeah, I mean, you can't see the other tabs in my browser right now, Alane, but it literally is this morning. You know, I just had this desire to essentially what you're saying, like, find specific solutions for specific problems that we're seeing right now. And I've got a whole bunch of different technology options open in front of me in tabs, and I'm working through which one makes sense, which one can a normal team actually implement.

0:29:02 - (Micah): Can you do this with easy integrations? Like, there's all this criteria these days where it's like, we don't want to pay development teams. We want to make this easy, we want to build easy. It goes all the way back to this whole upside down pyramid concept, which is, how do you operate at lightning speed? Because that's the only speed that you have access to.

0:29:24 - (Alane): Mm. All right, Micah, let's close this up. If anybody's listening and wants to chat more about one of the five things that we went through in the upside down pyramid or that we call the path to AI transformation, please reach out to us. Micah and I obviously have a passion about all five of these because we could have talked even longer about each one. Thanks for listening to this episode of Automate Your Agency. We hope you're inspired to take your business to the next level.

0:29:50 - (Alane): Don't forget to subscribe on your favorite podcast platform and leave us a review. Your feedback helps us improve and reach more listeners. If you're looking for more resources, visit our website at biggestgoal.ai for free content and tools for automating your business. Join us next week as we dive into more ways to automate and scale your business. Bye for now.