433-SYB-BTS-AI-Workflows.mp3: this mp3 audio file was automatically transcribed by Sonix with the best speech-to-text algorithms. This transcript may contain errors.
Carol Cox:
I’m taking you behind the scenes of how I’m integrating AI with real world use cases. On this episode of the Speaking Your Brand podcast. More and more women are making an impact by starting businesses, running for office, and speaking up for what matters. With my background as a TV political analyst, entrepreneur, and speaker. I interview a coach for purpose driven women to shape their brands, grow their companies, and become recognized as influencers in their field. This is speaking your brand, your place to learn how to persuasively communicate your message to your audience. Hi and welcome to Speaking Your Brand. I’m your host, Carol Cox. Today, I’m going to take you behind the scenes of how I’m using AI to power my business in new and innovative ways. Specifically, I’m going to show you AI automation workflows that I’m using to save time, create content, and scale impact. I’m hoping that this tool will inspire you to do something similar with your business in your content. So let me ask you this. If you could automate something that you do. What would that be? So think about all the things that you do in your business, whether they’re admin tasks, content creation, maybe even the way that you serve your clients. Is there something that you could automate that maybe didn’t seem possible before? But now with AI, specifically with AI agents, it could become possible. So keep that in mind as we’re going through today.
Carol Cox:
And by the way, if you’re listening to this on the Speaking Your Brand podcast, I am also recording this on video with slides. So if you would like to watch the video so you can see the slides in the demos that I’m going to be showing, you can get that on the show notes page for this episode as speaking your brand. 433. Today we’re going to talk about making the promise of AI a reality. I’m sure you hear a lot of conversations on AI online, and a lot of it is theory or what I would call wish casting. Like, what is it going to be like when I can do all of these things? But I really want to ground us, and I want to show you real world use cases of how it’s possible to use AI today. And we’re going to look at three areas of business operations, marketing and revenue. If you’re new to speaking your brand, welcome. I started speaking your brand ten years ago in 2015 to work with entrepreneurs and professionals on their public speaking and thought leadership. Before that, I was a software developer and I found it and ran two technology companies. We built large systems for fortune 500 companies like Office Depot and Lowe’s. So the tech part of me definitely came back to the surface when ChatGPT launched a few years ago and I immediately saw the potential of AI and how a kid transformed.
Carol Cox:
Not only what we do on the back end, but also in the way that we provide value to our clients in those technology projects that I did earlier in my career. We were doing a lot of automation and workflow efficiency, and I see AI as the next stage of automation, taking it even further than what was possible with programmatic deterministic software and now AI agents, which, given a goal, can go off and work on your behalf, whether it’s for minutes or even hours and potentially days or weeks in the not too distant future. Now, those represent a true paradigm shift when we think about what is possible for how work is going to be done, because the nature of work is changing for us. I really see AI as a collaborative partner, as even a creative partner, to allow us to do the more human work in allowing the AI and the robots to do the computer work. And as the ones who are guiding the AI. Our leadership and our management skills are going to be more important than ever. And what’s so mindblowing about the era that we’re in is that science fiction is quickly becoming reality. I think to the movie her, which came out in 2013, and in it, the main character played by Joaquin Phoenix, ends up falling in love with an AI named Samantha, who was voiced by Scarlett Johansson.
Carol Cox:
And when I initially watched this movie, I thought, there is no way this is ever going to be possible in my lifetime, maybe in some distant future, but certainly not while I’m alive. But we actually are in this era now. If you haven’t seen the movie or you haven’t seen it in a while, definitely go and watch it, because it’s uncanny how much the human and the AI have these very natural conversations with each other, very much like we’re having with ChatGPT and how Samantha, the AI can go into his computer and read his emails and summarize them and do work on his behalf. Well, that is where we’re at right now with AI agents, and they’re going to be even more and more sophisticated. Now, the challenge that many of us have in our businesses and as content creators is that there’s always so much to do. There’s always more content that we could create. And with how competitive the social media feeds and the algorithms are. If you feel like we have to create more and more content just to keep up, and of course, we also want to provide valuable content and services to our clients and to our audience. So I want you to think about for yourself and your team. If you have team members, are you spending too much time every day and every week on admin tasks, on content production, on business back end operations? And I want you to think that way, because those are the tasks that can be very easily done by AI.
Carol Cox:
Those are also the tasks that really should be done, in my opinion, by what I call the robots. Like the robots in the computer. And those are things like drafting emails, scheduling meetings, doing data entry, project management. Even some of the content production, the workflows related to content creation, social media market research, those types of things the AI is really good at. And by allowing the AI to do that, it frees up you and your team members to focus on more strategic activities like planning, business development, high touch client services, innovating within your business, public speaking, thought leadership, writing a book, being a guest on podcasts and in the media, and so on. Those are the things that you as the human should be doing, because only you can do it. After all, that is the promise of AI to allow us to automate, optimize, create and personalize in ways that we haven’t been able to before. So I’m going to walk you through some real world use cases and workflows that I’m using right now in speaking your brand to inspire you and to give you some ideas of what you can do as well. To help you do this, I’ve created a brand new live online program called Automate and Amplify with AI.
Carol Cox:
I’m going to show you, and we’re going to build together these AI automation workflows that you can use for your operations, your marketing, and your content creation. You get weekly zoom calls, a private slack group for Q&A and feedback. In between the zoom calls, you get personalized guidance and training from me, and I’m going to give you my automation workflow blueprints to get you started and show you how to customize them for yourself. You can get all the details of this program and apply as speaking your brand AI. Again, that’s speaking your brand. Am I? So here are the three areas that we’re going to look at for AI, automation, operations, marketing, and revenue. The first one is operations. And the first real world use case is creating an AI agent assistant. Now, just like in the movie her, I have had this dream where I would love to just be able to send a voice memo or even type something on my phone or on my laptop that says, can you please send an email and so and so and, and ask them if we can set up a meeting. And I don’t have to give the all the exact words and every sentence and every paragraph, I can just kind of give a general direction. And then the assistant knows what to do, or to schedule a meeting, or to add a task to my project management, or to take a lead and put it into asana, the project management tool I use and fill in all the fields.
Carol Cox:
Now, yes, I have had assistants in the past. Human assistants who can do this, but I want to free them up to do higher value activities. So the same thing with you. You may have a virtual assistant or an executive assistant who does these things for you right now. But imagine if you empower them to create these AI agent assistants and workflows. So then they can build those for you, maybe for other clients that they have to free them up to do those higher value human activities. For my AI agent assistant, I’m using slack as the interface. So I send a message either type it or via voice into slack. And then my workflow picks up that message and then does whatever it is that I’m asking it to do. Whether it’s draft an email, schedule a meeting, add something to asana, and so on. So I’m going to play the video demo here. If you are listening on the podcast, I’ll narrate what’s going on on the screen, but don’t forget that you can check out the video on the show notes page. Let me go ahead and play this demo. Send an email to Ed asking him to pack up the tripod and HDMI cable for our meetup presentation tonight.
Carol Cox:
Also, ask him if there’s anything he needs me to bring. All right, so I did that via voice. So I sent that message in asana. So now it’s going to my workflow and Macomb. And here’s what’s cool about this AI agent assistant is that it understands that it needs to send an email. So it knows is this is not adding to the calendar. It knows exactly what tool to go use, in this case Gmail. And it knows how to write the emails. So I don’t have to dictate the email word for word. Instead, I can just say what you heard me say. Send an email to editor, tell them to bring these things, see if there’s anything else, and then it’s OpenAI’s API ChatGPT behind the scenes, which is actually drafting that email. And with this AI agent assistant, I can keep thinking of different tasks, different things that I can incorporate. So more things that I can do on my behalf. Here’s the second use case in the operations category. There are now tools where you can create entire websites and entire web applications from a chat interface. So just like you chat with ChatGPT, you can do the same thing and build these entire websites and applications. It is amazing and it really allows you to innovate quickly. Say you have an idea for an app that you would want for your clients to be able to use, but it seems very daunting to have to figure out how to find a software developer, how to explain to them what you want, have it build, and then support it.
Carol Cox:
It’s it’s a lot. Trust me. As someone who did a lot of software development projects, it’s a heavy lift and there’s just a lot that goes into it. But now with tools like Replit from a chat interface, you can just ask it what to build. So I’m going to play this demo here, and I’m going to show you what it looks like. And what I had to do was create a website for an AI consulting service. So speaking, your brand AI is what I called it, and it built the entire website. It designed the entire website. And then in the chat interface I can then ask it to make modifications. So originally the website was when blue colors. So I told it. I said please change the colors to speaking your brand brand colors. And I gave it the purple and coral brand colors. So then it went through and it figured out where to change the colors. I didn’t have to tell it, change the color on the button, and then change the color on this heading, and then change this color on the gradient. It knew how to do that, and it will add graphics, or it will change out photos, and you just have to type it in the chat what you want it to do.
Carol Cox:
It is so cool. So those were for operations. So an assistant and being able to experiment and innovate with these tools like Replit to build websites and applications. Let’s look at the second category now which is marketing. So the first thing that I thought about is how to automate repetitive tasks. To save me time and to save my team time. We produce a podcast episode every single week, and there’s a lot that goes into it. Everything from editing the episode itself to writing the show notes, the LinkedIn post, the email newsletter, putting it on the blog, creating the episode graphics, and so on. I created an automation workflow using Macomb that goes through all of those different steps. All I have to do is take the final audio file, place it in a Google Drive folder, and that is it. And then the Macomb scenario sees that there’s a new file in that folder. And then kicks off the automation workflow to do the show notes, the LinkedIn post and so on. And then it runs through all of it. So all I had to do was drop it into that folder. So that not only saves time, but also just like that cognitive overload of having to go and remember to do all these things, or having to go do all these things.
Carol Cox:
And this is where automation is the next step from saying just using ChatGPT, using ChatGPT is amazing because then again, you can go, you can put a transcript in and it will write the show notes on everything, but it’s still a manual process. You’re still going to ChatGPT. Coming back out, you know, copying and pasting to a Google doc and coming back out. The one thing with using these automation workflows is that you do want to make sure that you’re providing very specific instructions to the large language model that you’re using. In this case, I’m using OpenAI’s API, which is ChatGPT. So I give it specific instructions to how to write the show notes and the email newsletter, and so on to make it sound like it’s coming from me. And speaking your brand with our brand voice and our style. And so in that automated amplify with AI program, I’m going to show you how to do this for yourself so that you make sure that that the output, the end result, sounds and feels like you. Not generic output. Now here’s the second real world use case in the marketing category was leveraging all of the content that I have by building a rag for my podcast. Now I’m going to I’ll tell you what a rag is in just a moment. I have over 400 episodes of the Speaking Your Brand podcast, which is a very rich content library.
Carol Cox:
But it also can be overwhelming for new listeners who find the podcast. They may have a specific question or a topic that they’re looking for, whether it’s speaking fees or finding events, or reducing nerves and building confidence or using stories in their presentations. And they have to scroll through a lot of episodes to find what they’re looking for. Also, I’ll feel like I’ve just done a topic such as storytelling, and I look back at my episode list and it’s been a year, maybe a year and a half since I dedicated an episode to it. So I wanted to find a way to make my podcast library much more accessible and usable to listeners. So that’s where a rag comes in. Now a rag is called a retrieval. Augmented generation engine is just a jargony way to say an AI knowledge base. So if you hear the term rag, just think it’s an AI knowledge base. So it’s like a knowledge base of all of your content, but AI is infused into it to make it much more powerful for you to access. And here’s the difference between just using a standard large language model versus using one with your rag with the standard large language model like ChatGPT. You ask it a question, or you enter a prompt and it’s going to give you a very well written response. But that response is based on all of its training data, basically in the entire internet.
Carol Cox:
So it’s going to be good, but it’s not specific to you your methodology, your processes, your frameworks, the way that you do things. That’s where a rag comes in. Because what happens is when you send that question or that prompt to the large language model, it looks at your rag, your AI knowledge base first to get the answer, and then it uses a large language model to write that very, very well-written output. So in this case, when it looks at my rag, it looks at my Speaking Your Brand podcast content to write the output. So now it mentions my frameworks and the way that we talk about doing things instead of just what’s on the internet. Here’s the workflow that I built for this. I use a company called needle Dicom for my Rag, my AI knowledge base, and I uploaded about 100 of my solo podcast episode transcripts to it. It’s sitting there in Needle eye. And then in my Macomb workflow scenario, if someone sends a question to podcast as speaking your brand, you could actually send a question. Their podcast is speaking your brand. It will look through my rag in needle to find the transcripts that have the pertinent content in it, and then open. I ChatGPT will write the response and then send the reply email to that person. So I’m going to show you here what it looks like.
Carol Cox:
So again I’m sending a question to podcast and speaking your brand. What is the signature talk and why would I need one. It’s going through the scenarios looking at my rag and needle ChatGPT is writing the response and then sending it back. And now here’s what’s so powerful about this is that the reply is based on how we as speaking your brand talk about signature talks. So it mentions our signature Talk Canvas framework. It talks about why you need a signature talk, and the way that we talk about it even can reference specific podcast episodes to point the person to. So not only is this really useful for podcast listeners to get on demand answers, but it also serves as brand awareness and lead generation for what we’re doing. So after I built this, I thought, well, text replies are nice, but what if people could hear my voice provide their reply? After all, a lot of these people are podcast listeners and they’re used to listening to me. So here’s real world use. Case number three is cloning my voice. I took about four hours of my podcast audio, and I used 11 labs to professionally clone my voice. Let me play you an example of an actual conversation between AI Carol, and a user who asked a question. Hi, this is AI Carol, from speaking your brand. How can I help you today?
Speaker2:
What’s the secret to creating a powerful signature talk?
Carol Cox:
Creating a powerful signature talk involves understanding your audience, having a clear message, and delivering it with authenticity and confidence. Is there anything specific you would like to know more about regarding signature talks?
Speaker2:
What if I don’t feel.
Carol Cox:
Okay, so then the conversation goes on. But again, I didn’t have to tell AI Carol, here’s exactly how to answer every single question that someone could ask instead, is looking at my podcast transcripts and figuring out the answer on the fly, and then having that very natural sounding conversation with the other person. You can try this out for yourself and speaking your brand. Com website. If you look in the lower right hand side of the website, you’ll see there’s a little pop up box that says have questions, ask AI Carol, and you can have either a text or a voice conversation. After I did this voice cloning, I started to think, well, if I can provide text replies, if someone emails a question, it could also provide voice replies as well. And this is truly personalization at scale, and I feel like a lot of content that’s going to be created with AI is going to allow this hyper personalization, this deep personalization based on what the person is looking for. So what I did in this case was very similar to sending an email asking a question, but instead of just getting back a text reply, now the person gets back a text reply and an audio file that has AI Carol, sharing a little bit more of the response to that question. You can try this out yourself by sending a question to On Demand as speaking your brand. And again, that’s on demand as speaking your brand. And so you’ll ask a question and then you’ll get back within a minute or so that reply with the audio message.
Carol Cox:
And then I started to think, well, these one off on demand replies are great, but what if I could create an entire companion podcast that is 100% AI generated with short answers to common questions that people have related to public speaking? Thought leadership, personal branding, business storytelling, and so on. So recently I launched the Confident Speaker podcast, which has episodes that are about 3 to 5 minutes in length and they’re all 100%. I created with my voice clone. I have a list of topics in a spreadsheet, and then it goes through and creates an episode based on the next question or the next topic in the spreadsheet. So it’s using my AI knowledge base, all of those podcast transcripts. Openai is writing the script and the show notes the workflow automation, and Macomb does all of this sends the script to 11 labs, my voice clone to actually generate the audio and then ultimately sends it to my podcast host, transistor. It is in draft mode, so I can go check the script and show notes and listen to the audio, and it takes me longer to do that than it takes to create the episode. It really takes about one minute for the entire workflow to run to create an episode. Now contrast that with about the four hours on average it takes for me to create a regular Speaking Your Brand podcast. I’m excited about this because it’s a new content channel that I’ve created. It allows people to get quick answers to what they’re looking for, and builds more brand awareness for speaking your brand and hopefully lead generation as well.
Carol Cox:
Let me play a clip from one of those AI generated episodes. Welcome to the Speaking Your Brand podcast. This is your host AI Carol, have you ever looked out at your audience mid speech and thought, are they even still with me? We’ve all been there talking on a stage, making our points only to notice glazed eyes, distracted glances, or the worst, someone pulling out their phone. Now, would you have known that that was AI Carol, and not real Carol? Probably not. That’s how good this voice cloning is. Now some people. Now some people have asked me, well, aren’t you concerned that people could take your voice and use it? I personally am not concerned. I’m not famous or a celebrity. Plus, there’s hours and hours of my voice already out on the internet from all of my podcast episodes. And with 11 labs. After I uploaded the four hours of podcast audio, I had to record into the software live. Me saying a paragraph so it could match that podcast audio to me, the real person, to make sure that I wasn’t trying to clone someone’s voice, that I didn’t have permission to do so. So there are safeguards built in. Now, of course, is there software out there that people could do this without someone’s permission? Yes. But again, I’m not worried about that. I’m not well known enough to do that, but what it has allowed me to do is to create content and create value to my audience without me having to spend hours doing so.
Carol Cox:
And by freeing up that time, I can now host more workshops, whether online or in person. I can do more one on one work with clients. I can run that new automated amplify with AI program because I’m not spending time doing these repetitive tasks. Let’s take a look at that third area of opportunity, which is revenue. Thinking about how to integrate AI into, say, the client work that you’re doing. The first real world use case under revenue is lead nurturing. So think about lead nurturing and business development. And this is where you can really use personalization at scale. One of the things that is always on my weekly task list that honestly I never get to, is finding new connections on LinkedIn and I. And what I like to do is when someone signs up for our email list and ConvertKit is go look them up on LinkedIn, read a little bit about them, and then send them a connection request with a nice message. And I don’t really get around to doing that, even though I wished I did. I built an automation in Make.com, so when someone subscribed to the email list and ConvertKit, it sends it over to Airtable. My spreadsheet perplexity I then finds their LinkedIn profile, summarizes it, looks to see some suggested speaking topics, and then writes a personalized LinkedIn message to the person based on the work that they do and any commonalities between what they do and what we do.
Carol Cox:
As speaking, your brand suggested speaking topics for them and so on. And of course, I look at the message before I send it to the person, but this saves me easily 10 to 15 minutes per person. This is the type of activity that I wanted to be doing, but just never got around to doing it. And now this is the type of activity that I can do really well. The second real world use case under revenue is thinking about creating additional services for your clients. In our case, we’re creating a DIY service for our signature talk process. I’m calling it chat CIB, so like ChatGPT said, this chat speaking your brand chat CIB, we’re basing it on our signature Talk Canvas framework, which we’ve used with hundreds of clients over the years and is a proven framework to create a compelling and engaging talk that provides transformation, not just information to your audience. You primarily do this via a one on one service with our clients, so we do a three hour VIP day, primarily on zoom, sometimes in person, where we’re asking the client a bunch of questions about the work that they do on their message, and then we’re mapping it out on our poster board with the different posts and notes. And it’s an extremely effective process, but it is time consuming and it is a high touch premium service. So I’ve been thinking, how can we provide a DIY version for people to use as well? So that’s where chat CIB comes in.
Carol Cox:
And I mentioned Replit earlier that allows you to build websites and web applications, where I’m using Replit right now to build a prototype of Chat Sibi, and it is amazing because it allows me to get all of my ideas out really quickly and see them come to life. The third real world use case and their revenue is to share and monetize your knowledge. And in my case, I’m doing this with my new Automate and Amplify with AI program, where I want to share with you what I’ve built and what I’m going to be continuing to build, so that you can use these workflows and these automations in your own business and with your own content creation, to free up your time to do more of the human work that you would like to be doing. And don’t forget, you can get all the details and apply for this program at Speaking Your Brand. So here are the next steps I want you to think about using AI not only as a collaborative partner, but really to innovate. To innovate what you’re doing in your business and with your content creation. And last week’s podcast episode 432. I shared four questions from Wharton professor Ethan Morlock about how to think about AI, specifically in a business context. And the four questions are. Number one, what thing do you currently do is no longer useful, meaning no longer useful in your work or to your clients because I can do it better and faster.
Carol Cox:
Number two, what impossible thing is now possible for you to do? And in my case, it’s creating that 100% AI generated companion podcast. Number three, what can you democratize access to in the work that you do? So chat CIB, that DIY service is an example of us democratizing our services. And number four, what can you personalize. So how can you personalize more what you’re doing with AI. So that’s how I want you to think about this idea of collaboration and innovation with AI and with automation workflows. So think about for your business for operations, marketing and revenue. What are regular repetitive tasks that can be automated? What are activities you’re not currently doing but could or should be doing that you now have time for? Or that could be more automated like that. Lead nurturing with LinkedIn that I mention. And how can you use AI to better understand your content and your marketing so that you can serve your team and your clients and audience in a better and new way? I invite you to shift your mindset to allow AI to do the admin and marketing tasks, so you and your team can focus on the strategic and the human activities. In addition to that, automate and Amplify with AI program. Here at Speaking Your Brand, we provide coaching and training for both individuals and teams on public speaking, executive presence, thought leadership, and business storytelling. You can get details about our programs and workshops as speaking your brand. Com. Until next time, thanks for listening.
Sonix has many features that you’d love including powerful integrations and APIs, automated translation, automated subtitles, advanced search, and easily transcribe your Zoom meetings. Try Sonix for free today.
