Show Notes
Scott and his team are using AI agents across multiple businesses. Here are the nine use cases:
1. Data Analysis: AI agents pull data from multiple sources, calculate metrics, identify trends. Most small businesses don't have data analysts. Now they can.
2. Revenue/Financial Analyst: Daily revenue calculations, transaction reconciliation, refund trend analysis, investment return evaluation. Paired with a human who reviews daily.
3. Business Operations Assistant: Reduces operational load. Example: finishing a studio recording kicks off file downloads, editor notifications, and workflow—all triggered by telling the agent "this is done."
4. Billing Investigations: AI reviews billing records and catches human errors. Found thousands of dollars in improper billing. Revenue recovery.
5. Automation Builder: Builds internal tools and scripts. Example: studio background auto-switches based on what Scott is recording.
6. Automation Supervisor: Daily 6 PM review of all automations. If something broke and didn't self-correct, the supervisor investigates, fixes it, and reruns it. Prompt: "You work for a self-improving organization. When something fails, it is your job to fix it."
7. Research/Due Diligence: Market research, property due diligence, historical records. "You'd be amazed at how much due diligence can be done by AI."
8. Customer Support (Tier 1): Drafts responses within defined rules. 90% confidence threshold—below that, escalate to a human. Never sends without review.
9. Customer Analysis: Before any call or email, the agent examines the full customer picture—email history, billing, meeting notes, last human touchpoint. Enables deeper conversations.
The $80K story: An employee left. Management said: "Give us $20K for AI tools and we can augment her work." It's happening.
The principle: Every use case pairs AI with a human. "It's never the AI running rampant. It's a true partnership."
Got a business question? Ask Scott here: scotttodd.net/ask
📜 Full Transcript (Click to expand)
Scott Todd (00:02.248)
Last week we talked about AI agents. I talked about how twenty twenty six is the year I of AI agents for me and for my team and how we're using, you know, really every framework that came down this year from open claw back in January to Hermes to Buzz to now Grockbot. And really in my company, we've kind of narrowed it down into really using Hermes and also Grokbot. And what
We're I I like both of them. Okay, like I think that they each have their own different places within our organization. And in this episode, what I thought I would do is share with you some of the broad use cases that, well, me and my team are using and for these AI agents and how we're using it and kind of give you maybe some ideas of how you might think about using it within your own organization. Welcome to Fix My Business. I'm your host, Scott Todd. I have built multiple seven-figure businesses. After leaving my
Corporate fortune three hundred VP job. I did that ten years ago, and my whole goal with this channel is to help you build a business that you love and that will work for you. So let's look at these nine use cases. And again, this isn't just me and how I'm using it. It's also how my team is using it to help our organizations. So here's the th first part is
Our bots, our AI agents, they are really, really, really embedded in our data analysis. They serve as probably some of the best data analysts that, well, I never really had. Look, most businesses, small businesses at least, do not have data analysts that are looking at all of these different metrics and all of these different financial numbers. And with AI and AI agents, we can do that. We can hire these data analysts or data agents that basically
allow us to pull data from multiple sources, calculate performance metrics, identify trends that, well, basically most small business owners are too busy to ignore. Okay, let's just be honest. Most business owners do not think about their trends like larger companies do. And really to explain what the numbers mean. And I think that that's one of the big disadvantages that small businesses encounter when they compare to large businesses is really this data
Scott Todd (02:28.23)
analysis team. Large companies have large peop large groups of people that analyze data. And look, it's not cheap. These financial analysts could easily cost eighty thousand dollars right out of college. Okay, to put that into a smaller organization, it really doesn't happen. So it's left to well the business owner and also the team that's there to really take time to dig into the numbers. But what we can do is we can now turn this AI agent loose with our data
And bring us back trends. Hey, what are we missing? How can we not how can we think about things in different ways? A few weeks ago I talked about churn and about how in one of my organizations of software, we have this metric where we look at churn and ways that we can dig deeper into the churn. If they met this requirement, then that their likelihood of churn was less. If they met that requirement, it was less. And it's helped us to transform the way that we see the customer journey because now we can.
begin to get ahead of this and say, hey, wait a minute. This is a potential risk of losing a customer. And when we have that data, now we can take action on it. And it's something that we'll take the data and give it to a team member who can now react to it. And we can kind of figure out how we can use the data to get better within our organization. The second way that we're using this like never before is a revenue and financial
analyst and what I mean by that is every day we are obviously doing the simple things like calculating revenue, reconcile transactions, but this is all something that someone else had to do. a human had to do. But now we're able to pair this analyst, this AI agent with a human. I think that's a key word I want to stress, we pair it with a human that will go out and look every day
at this large horizon and say, hey, what are we missing? And here's some trends that we we really don't have. So think about the data analyst. It's different than the financial analyst because the financial analyst is all about finances. Okay? Now these AI agents can also investigate, I don't know, refunds, look for trends for refunds. We can evaluate investment returns. We can look at our business economics in a way that, well, we could really never do without
Scott Todd (04:54.233)
Either more people or more resources. And that's where the AI agents come in. The third way is our business operations assistant. So for example, I have an agent that I will give an operational outcome to. Hey, I want you to do this thing. I want you to correct this record. I want you to process this report. I want you to update this account. I want you to transfer these files. I want you to to process these recurring workflows.
And it all is designed to to basically reduce the operational load. Let me give you example. So when I do I do a weekly video for YouTube on a different channel for a different business, when I do that and I complete it, it kicks off a series of workflows. It goes to a video editor, it goes gets har downloaded from our studio that's here into our hard drive storage.
All of this has to happen. And in the past, it required me to go to a computer and say, okay, this is done. Let me download these files and kick off the workflow. Now I just say to the AI agent, hey, this is done. It kicks off the first process, which is to move move it down. It kicks off recurring workflows so that the team is notified. And basically it it reduced my operational requirement dramatically.
The fourth way that we're using AI agents is for billing and account investigations. So what we're doing is we're actually having AI within our billing records, within our data. And when we have complicated account histories or we're trying to understand what happened with a bill or an invoice, we no longer have to have a human go do this. And what happens is that now that the AI agent can look at this.
What we're finding is we're finding revenue leakage where people weren't billed correctly. That's a problem. And it's the AI agent is saying, Hey, I found this, I found that. Just the other day, it found multiple hundreds of dollars of just an improper billing. A human made a mistake. That's the way that it worked. The AI agent caught it. There's been multiple cases, thousands of dollars, where the AI agent determined, like, hey,
Scott Todd (07:19.319)
This was not set up correctly and someone hasn't been paying the right amount. That was a human error that the AI agent caught. So now we're able to basically shure up our human basically errors and go back and we're we're driving more revenue because people are paying what they should pay. The fifth way that we're using AI within the organization is as an automation builder. And this is probably one of the easiest no-brainers.
That most people jump into. And there was so much that we continue to automate within our business. We were using all these different tools. You probably use them as well. And we've built a lot of internal tools with AI. And as a result of that, we've been able to script certain things from happening. So let me give you just an automation example. If you look at if you're watching this on video and you look behind me, there's a basically a monitor. It has a graphic on there.
In order to set that up, my team would need to know that, hey, Scott was going to be in the studio on this day to record this at this time. And the team would go in and they would essentially change this background. It would change it to this background, or if I was in the studio for something else, it would change it to the next background. But a team member had to go do that. And then what happened was we were able to t basically tell our age AI agent, our automation agent, hey,
Scott's going in the studio now, change the background. And by the time I get in here, the background is set to be what it's meant to be. So it's just a human that does not have to go and do a task, a computer-based task. The agent can do that. The sixth way that we're using it is as an automation supervisor. So look, one of the one of the biggest realities of automation is that automation's
Obviously, either succeed in every run or they fail. And too many times when we build automations, it's easy to forget about the notification. You either get the happy path, like, hey, the report is just there, and you see it because you go use this report every day or this output every day. But then if you're not using this report every single day, but you're counting on it being there, and then when you go need to go use it, it's not there. Guess what? You find out that the automation broke. Now
Scott Todd (09:43.454)
we can talk about the fact that the automation wasn't built correctly. I agree. We could talk about the fact that we should have notifications within the automations. I 100% agree. In fact, if you go back into this channel's archives or to the to the podcast archives, you'll see many times where I talk about the fact that we need to build with scale. And one of the the the the L in scale is for leverage the data, which means that we need to notify be notified when it works and when it doesn't work.
But now we're able to use an AI supervisor, an AI automation supervisor that looks for this, that it's looking for the failed task. And in my organization with our AI agents, the thing that we tell them is, hey, you work for a self improving organization. Meaning that when something fails, it is your job to fix it. That's a prompt that we have in all of our automation supervisors kind of you know.
ex explanation or history of what they're supposed to do. Self-improving organization. If something is broken, it is your duty to fix it. And I'll tell you, we also run a daily script. It's at 6 PM Eastern Time, where we do a r daily review of all the automations. If something failed to work that day, if a job failed to run that day, if it didn't work and it didn't correct itself, this AI supervisor or automation supervisor will go in there.
And figure out what happened, how do we fix it, fix it and run it at that time so that we're not missing automations. It's somewhat simple, but it's something that you should really think about. The seventh way is as a research and due diligence assistant. So what I might have it do is I might have it research markets for me or different businesses for me, or go look at, you know, different
historical records of what it can uncover and find, whether it's it's internal or external on the web. Anything that requires research, we use these as research assistants. And including in my investing business, we might say, hey, go run due diligence on this particular property. Here's what the criteria for success looks like. And you will be amazed at how much of due diligence can be done by AI assistants. It's
Scott Todd (12:09.265)
Mind boggling. The eighth way is really customer support. And here's the thing is with our customer support, it is tier one support, meaning we've defined rules, we've defined what success looks like. We take customer questions, it will draft, it will draft a response. It does not send the responses unless it's w within the required defined rules.
So one of the big things that we have there is if you're not ninety percent sure that this is the right answer, then you escalate it to a human. Again, what we're trying to do with all of our AI agents is we're trying to pair them up with a human so that it's never the AI that is just running rampant and controlling the business. It is a it is a true partnership, human and AI. And it will make you
more productive and it's made my team way more productive. And in fact, as some employees have left, we the the team has come to me and said we don't need to replace them because we can if you just give us, you know, a a portion of what their salary was, we can replace them with AI. And in fact, I was talking to a business owner last week or maybe it was two weeks ago and he told me that an eighty thousand dollar a year employee left this company
And the leadership team came to him and said, Hey, so-and-so is leaving. She was a fantastic employee. And he was like, Wow, I'm really bummed. She's been with us for a very long time. Wow, what are we gonna do? And the management team, his management team said to him, Hey, she was making eighty, but if you give us a budget of twenty thousand dollars, we believe that we can leverage AI to augment her work and the team can help augment that.
Again, I I hate hearing that people are not are are being replaced by AI, but at the end of the day, this is where efficiencies are going to come from, is our teams need to be able to gain efficiencies from the AI. And that includes that includes, you know, people that leave, and we need to allocate some portion of their pay to an AI budget to offset that. And then finally, I would say that it's comes back down the ninth.
Scott Todd (14:37.038)
business case is really customer analysis. So what happens is we might be dealing with a customer and before we make a phone call, before we get on a call with them, before we exchange an email, we want to h examine the whole picture. What what how have they interacted with our ecosystem? So we're able to do that because the AI agent has access to every single touch point that they may have come through, including email, including
you know, calendar invites, calendarly or meeting notes that we have or billing systems. We wanna understand what this entire picture looks like. When was the last time a human within the organization touched their account, looked at it? What was what was entailed in there? And as a result of that, what we're able to do is we're having able to have deeper conversations and deeper understanding of where customers might be struggling. And that also allows our customer success team
A different way of looking at the customer accounts. So that's nine ways in which we're using AI agents within my organization. I hope that you have some new ideas there on how you can implement these into your organization. And next time, what I'd like to talk about is really something I mentioned at the beginning of this call, which was this: Hermes agent, they do certain things. And I find that GrockBot agents.
are better at other things. And so oftentimes people will say, which one should I use? And I say use as many of them as you can afford. Because my Grokbot agents, they do one set of skills, whereas my Hermes agents do something that's completely different. And yet somehow they communicate and our teams continue to leverage both different frameworks. So that's what we'll talk about next time.
If you have a business question, if you want to chime into this, please be sure to visit scottodd.net forward slash ask. And I will see you in our next episode.
Last week we talked about AI agents. I talked about how twenty twenty six is the year I of AI agents for me and for my team and how we're using, you know, really every framework that came down this year from open claw back in January to Hermes to Buzz to now Grockbot. And really in my company, we've kind of narrowed it down into really using Hermes and also Grokbot. And what
We're I I like both of them. Okay, like I think that they each have their own different places within our organization. And in this episode, what I thought I would do is share with you some of the broad use cases that, well, me and my team are using and for these AI agents and how we're using it and kind of give you maybe some ideas of how you might think about using it within your own organization. Welcome to Fix My Business. I'm your host, Scott Todd. I have built multiple seven-figure businesses. After leaving my
Corporate fortune three hundred VP job. I did that ten years ago, and my whole goal with this channel is to help you build a business that you love and that will work for you. So let's look at these nine use cases. And again, this isn't just me and how I'm using it. It's also how my team is using it to help our organizations. So here's the th first part is
Our bots, our AI agents, they are really, really, really embedded in our data analysis. They serve as probably some of the best data analysts that, well, I never really had. Look, most businesses, small businesses at least, do not have data analysts that are looking at all of these different metrics and all of these different financial numbers. And with AI and AI agents, we can do that. We can hire these data analysts or data agents that basically
allow us to pull data from multiple sources, calculate performance metrics, identify trends that, well, basically most small business owners are too busy to ignore. Okay, let's just be honest. Most business owners do not think about their trends like larger companies do. And really to explain what the numbers mean. And I think that that's one of the big disadvantages that small businesses encounter when they compare to large businesses is really this data
Scott Todd (02:28.23)
analysis team. Large companies have large peop large groups of people that analyze data. And look, it's not cheap. These financial analysts could easily cost eighty thousand dollars right out of college. Okay, to put that into a smaller organization, it really doesn't happen. So it's left to well the business owner and also the team that's there to really take time to dig into the numbers. But what we can do is we can now turn this AI agent loose with our data
And bring us back trends. Hey, what are we missing? How can we not how can we think about things in different ways? A few weeks ago I talked about churn and about how in one of my organizations of software, we have this metric where we look at churn and ways that we can dig deeper into the churn. If they met this requirement, then that their likelihood of churn was less. If they met that requirement, it was less. And it's helped us to transform the way that we see the customer journey because now we can.
begin to get ahead of this and say, hey, wait a minute. This is a potential risk of losing a customer. And when we have that data, now we can take action on it. And it's something that we'll take the data and give it to a team member who can now react to it. And we can kind of figure out how we can use the data to get better within our organization. The second way that we're using this like never before is a revenue and financial
analyst and what I mean by that is every day we are obviously doing the simple things like calculating revenue, reconcile transactions, but this is all something that someone else had to do. a human had to do. But now we're able to pair this analyst, this AI agent with a human. I think that's a key word I want to stress, we pair it with a human that will go out and look every day
at this large horizon and say, hey, what are we missing? And here's some trends that we we really don't have. So think about the data analyst. It's different than the financial analyst because the financial analyst is all about finances. Okay? Now these AI agents can also investigate, I don't know, refunds, look for trends for refunds. We can evaluate investment returns. We can look at our business economics in a way that, well, we could really never do without
Scott Todd (04:54.233)
Either more people or more resources. And that's where the AI agents come in. The third way is our business operations assistant. So for example, I have an agent that I will give an operational outcome to. Hey, I want you to do this thing. I want you to correct this record. I want you to process this report. I want you to update this account. I want you to transfer these files. I want you to to process these recurring workflows.
And it all is designed to to basically reduce the operational load. Let me give you example. So when I do I do a weekly video for YouTube on a different channel for a different business, when I do that and I complete it, it kicks off a series of workflows. It goes to a video editor, it goes gets har downloaded from our studio that's here into our hard drive storage.
All of this has to happen. And in the past, it required me to go to a computer and say, okay, this is done. Let me download these files and kick off the workflow. Now I just say to the AI agent, hey, this is done. It kicks off the first process, which is to move move it down. It kicks off recurring workflows so that the team is notified. And basically it it reduced my operational requirement dramatically.
The fourth way that we're using AI agents is for billing and account investigations. So what we're doing is we're actually having AI within our billing records, within our data. And when we have complicated account histories or we're trying to understand what happened with a bill or an invoice, we no longer have to have a human go do this. And what happens is that now that the AI agent can look at this.
What we're finding is we're finding revenue leakage where people weren't billed correctly. That's a problem. And it's the AI agent is saying, Hey, I found this, I found that. Just the other day, it found multiple hundreds of dollars of just an improper billing. A human made a mistake. That's the way that it worked. The AI agent caught it. There's been multiple cases, thousands of dollars, where the AI agent determined, like, hey,
Scott Todd (07:19.319)
This was not set up correctly and someone hasn't been paying the right amount. That was a human error that the AI agent caught. So now we're able to basically shure up our human basically errors and go back and we're we're driving more revenue because people are paying what they should pay. The fifth way that we're using AI within the organization is as an automation builder. And this is probably one of the easiest no-brainers.
That most people jump into. And there was so much that we continue to automate within our business. We were using all these different tools. You probably use them as well. And we've built a lot of internal tools with AI. And as a result of that, we've been able to script certain things from happening. So let me give you just an automation example. If you look at if you're watching this on video and you look behind me, there's a basically a monitor. It has a graphic on there.
In order to set that up, my team would need to know that, hey, Scott was going to be in the studio on this day to record this at this time. And the team would go in and they would essentially change this background. It would change it to this background, or if I was in the studio for something else, it would change it to the next background. But a team member had to go do that. And then what happened was we were able to t basically tell our age AI agent, our automation agent, hey,
Scott's going in the studio now, change the background. And by the time I get in here, the background is set to be what it's meant to be. So it's just a human that does not have to go and do a task, a computer-based task. The agent can do that. The sixth way that we're using it is as an automation supervisor. So look, one of the one of the biggest realities of automation is that automation's
Obviously, either succeed in every run or they fail. And too many times when we build automations, it's easy to forget about the notification. You either get the happy path, like, hey, the report is just there, and you see it because you go use this report every day or this output every day. But then if you're not using this report every single day, but you're counting on it being there, and then when you go need to go use it, it's not there. Guess what? You find out that the automation broke. Now
Scott Todd (09:43.454)
we can talk about the fact that the automation wasn't built correctly. I agree. We could talk about the fact that we should have notifications within the automations. I 100% agree. In fact, if you go back into this channel's archives or to the to the podcast archives, you'll see many times where I talk about the fact that we need to build with scale. And one of the the the the L in scale is for leverage the data, which means that we need to notify be notified when it works and when it doesn't work.
But now we're able to use an AI supervisor, an AI automation supervisor that looks for this, that it's looking for the failed task. And in my organization with our AI agents, the thing that we tell them is, hey, you work for a self improving organization. Meaning that when something fails, it is your job to fix it. That's a prompt that we have in all of our automation supervisors kind of you know.
ex explanation or history of what they're supposed to do. Self-improving organization. If something is broken, it is your duty to fix it. And I'll tell you, we also run a daily script. It's at 6 PM Eastern Time, where we do a r daily review of all the automations. If something failed to work that day, if a job failed to run that day, if it didn't work and it didn't correct itself, this AI supervisor or automation supervisor will go in there.
And figure out what happened, how do we fix it, fix it and run it at that time so that we're not missing automations. It's somewhat simple, but it's something that you should really think about. The seventh way is as a research and due diligence assistant. So what I might have it do is I might have it research markets for me or different businesses for me, or go look at, you know, different
historical records of what it can uncover and find, whether it's it's internal or external on the web. Anything that requires research, we use these as research assistants. And including in my investing business, we might say, hey, go run due diligence on this particular property. Here's what the criteria for success looks like. And you will be amazed at how much of due diligence can be done by AI assistants. It's
Scott Todd (12:09.265)
Mind boggling. The eighth way is really customer support. And here's the thing is with our customer support, it is tier one support, meaning we've defined rules, we've defined what success looks like. We take customer questions, it will draft, it will draft a response. It does not send the responses unless it's w within the required defined rules.
So one of the big things that we have there is if you're not ninety percent sure that this is the right answer, then you escalate it to a human. Again, what we're trying to do with all of our AI agents is we're trying to pair them up with a human so that it's never the AI that is just running rampant and controlling the business. It is a it is a true partnership, human and AI. And it will make you
more productive and it's made my team way more productive. And in fact, as some employees have left, we the the team has come to me and said we don't need to replace them because we can if you just give us, you know, a a portion of what their salary was, we can replace them with AI. And in fact, I was talking to a business owner last week or maybe it was two weeks ago and he told me that an eighty thousand dollar a year employee left this company
And the leadership team came to him and said, Hey, so-and-so is leaving. She was a fantastic employee. And he was like, Wow, I'm really bummed. She's been with us for a very long time. Wow, what are we gonna do? And the management team, his management team said to him, Hey, she was making eighty, but if you give us a budget of twenty thousand dollars, we believe that we can leverage AI to augment her work and the team can help augment that.
Again, I I hate hearing that people are not are are being replaced by AI, but at the end of the day, this is where efficiencies are going to come from, is our teams need to be able to gain efficiencies from the AI. And that includes that includes, you know, people that leave, and we need to allocate some portion of their pay to an AI budget to offset that. And then finally, I would say that it's comes back down the ninth.
Scott Todd (14:37.038)
business case is really customer analysis. So what happens is we might be dealing with a customer and before we make a phone call, before we get on a call with them, before we exchange an email, we want to h examine the whole picture. What what how have they interacted with our ecosystem? So we're able to do that because the AI agent has access to every single touch point that they may have come through, including email, including
you know, calendar invites, calendarly or meeting notes that we have or billing systems. We wanna understand what this entire picture looks like. When was the last time a human within the organization touched their account, looked at it? What was what was entailed in there? And as a result of that, what we're able to do is we're having able to have deeper conversations and deeper understanding of where customers might be struggling. And that also allows our customer success team
A different way of looking at the customer accounts. So that's nine ways in which we're using AI agents within my organization. I hope that you have some new ideas there on how you can implement these into your organization. And next time, what I'd like to talk about is really something I mentioned at the beginning of this call, which was this: Hermes agent, they do certain things. And I find that GrockBot agents.
are better at other things. And so oftentimes people will say, which one should I use? And I say use as many of them as you can afford. Because my Grokbot agents, they do one set of skills, whereas my Hermes agents do something that's completely different. And yet somehow they communicate and our teams continue to leverage both different frameworks. So that's what we'll talk about next time.
If you have a business question, if you want to chime into this, please be sure to visit scottodd.net forward slash ask. And I will see you in our next episode.