Unknown

0 views10268 WordsCopy TextShare
Unknown
Video Transcript:
Okay, we are live. So welcome everyone to today's webinar about the topic of 10x podcast and webinar content with AI. I'm Tham and together with me today is Tuan here we are the founders of MindPal and we would be very excited to walk you through how exactly our customers and we ourselves have been leveraging AI for podcasts and webinar content marketing. So for today's webinar, this would be our agenda. We would first briefly talk about why this webinar in the first place. My co founder will share his insights about how he's been observing different use
cases of AI in business and why we decided to host a dedicated webinar about AI use cases specifically for podcasts and webinar content. Secondly, we will look at 6 real life use cases about how you can use MindPal in the context of podcasts and webinar content. These use cases are real and actual use cases that we have implemented ourselves in our business and also what we have been observing and assisting our customers to to do. So we hope that you would find our breakdown really helpful today and we will also look at some best practices
of how you can implement these AI use cases and then we would have a little bit of time for Q and A. So if you have any questions, I know we have some viewers from our YouTube live streaming and also some participants in our Zoom meeting. So feel free to drop your questions inside the chat or, or inside on YouTube or on Zoom and our team will answer them by the end of this webinar. So without further ado, let's get straight into it. First, I would like to invite Duong, my co founder to share about
why we decided to host this webinar. So why podcasts and webinar content specifically? Can you share a little bit? Dong? So we have been organizing and record quite some webinar about content creation or some of the about SEO about how to do different stuff with AI workflow in the business. And as we observe, a lot of customer are using MindPal and using specifically using the motivation workflow. I think that podcast and webinar content is a mixed part of the usage right now because you know, now a lot of people doing podcasts as maybe a side
hustle projects or serious project and a lot of company including MindPal using webinar as a way to engage with customer and as a way to finding a new customer and building the relationship and even closing sales in a webinar. So that is a big part of the business and there is a Growing number of people doing podcasts right now. But the process to do the podcast is not just about doing the podcast. You need to prepare for the podcast. For example, you need to do some research about the guest speaker. You need to deep dive
into some of the article about the topic, because you might do not know a lot of things. For example, if we do a podcast about open source AI model right now, we need to do a lot of research about how this model are doing on the market, for example, or if we interview someone, we may need to do research about them. And we need to prepare for the podcast, the podcast outline, the transcript, the slide. And that is a before podcast, but after podcast, we need to distribute it. We need to create content on social
media to promote the podcast. We build a new lecture after the podcast to summarize some insight to our audience. And that is a workflow that will require, I think maybe 5 to 10 hours of work for each podcast and even more with some of the more sophisticated podcast and webinar content with a lot of customers. And even for the case of webinar, for example, when we, when I observe some of the company, they organize the webinar, they even have the email campaign before the webinar that have several emails to remind people. And building those stuff
is actually not difficult. It will follow some template and have some, if we have some business contact after that, if you have some templates template to do it. And we can totally automate that part with AI. So instead of spending a few hours to craft by hands those email or those repurposed content, we can set up a workflow in maybe 30 minutes. But then after that, for every single webinar, we only take around 5 to 10 minutes to prepare everything. And we will save a ton of time. So I think this is a topic that
deserves a dedicated section from our team to do that. Yes, I totally agree. And I also speak from my own experience. You guys probably know that recently we have been hosting more webinars to discuss different use cases of MindPal. We have found that to be a very great way to connect with our customers. But at the same time, I think, I think for me, as the key host of these webinars, I find out that the most rewarding part of this would be the actual webinar. When I got the chance to deliver my insights and my
experience to the audience. But all the other preparation work that goes before that or after that, it could be really daunting because I would need to like prepare the event titles, brainstorming Event titles, how to make it more compelling, writing the event descriptions, writing these email to introduce the event to my audience, the social media posts, all sorts of things and that is. That's not to mention there are a bunch of different things I would need to do afterwards too because each webinar like this is a treasure trove of insights and content. Because you know,
speaking is the fastest way for us to produce content and by directly interact with the audience we can actually produce a lot more insights and content ideas that are not are not possible if we just think about them by ourselves. So I really want to maximize all of these insights and content that I have produced during the webinar to ensure that we get the highest ROI for this activity. Because of course this is a very time consuming task. So yeah, so that's why we after a while of doing this ourselves and also helping some customers,
implementing all sorts of AI agents and multi agent workflows for podcasts and webinar content, we decided to consolidate all of our learnings and observations into this one webinar so you can get more ideas to apply to your own business as well. So with that in with the context set already by Dong, we can go straight into 6 real life use cases of my powerful podcast and webinar content. And we have divided this into two main categories, pre event and post event. So the each business may have your own way of preparing and distributing this content,
but we believe there are a lot of patterns shared by all of us. So we have built some AI workflow templates that are ready to be shared to all of you so you guys can take them and customize them for your own business. Those six use cases include one of first of all is a comprehensive event planner that covers everything from brainstorming titles, writing event description, writing promotional posts, emails, sequence, et cetera for you. Then we will look at an AI workflow for guest researcher actually because based on our experience there are a lot of
podcasts and webinars done by our customers that involve the like the that has a guest speaker or many guest speakers and they often need to do a lot of research before that to make sure the podcast and webinar the most meaningful and fruitful as possible. And we will also take a look at a content prep AI workflow which is. Which is something I find personally very useful which will help you kind of brainstorming and planning for the content for your next podcast and webinar. We will get to that in a moment. Then we will take
a look at some post event Workflows that we can use MindPal for. These include comprehensive event recap content generator workflow. It can generate all content for you. This one is very comprehensive, I assure you. It has something like nearly 20 steps and it's. And we will take a closer look at that later on. But just a sneak peek, I think this one is the most impressive AI workflow I have ever seen. And next we will also look at an AI workflow for event quality review. So something we also use for ourselves too because. But I
need to admit that I'm not a professional presenter. So I need to constantly review my own performance as the host to identify areas for improvement and how to improve for my next next webinars. You know our first webinar, we don't. We didn't even have a slide like this. So this AI workflows has helped me to get more ideas on how I can improve to make sure the audience would get like best experience watching the webinar too. And then we have digital material Jetbot, which is a very interesting idea I learned from some of my customers.
They essentially they built a chatbot that learned from all of these webinar recordings and are able to answer the participants questions related to the webinar content. So to help them navigate their library of constantly growing webinar recordings and digital materials, let's get into them these AI workflows and chatbots, shall we? Now let me. Can you guys see my screen here? You can see my screen, right? I'm sharing the MindPal workspace. No, I still see the slide. Okay, let me share it again. Perfect. First, let's start with three AI workflows for pre event processes. The first
one as I have shared is this end to end and event planner workflow. This is based on my own experience preparing webinars like this too. For every webinar I noticed that my workflow typically involves after I discuss with my team about what is the brief of the next webinar, what is the purpose, who is the target audience, some logistics details like time and location and the format and who is the host. Do we have any guest speakers, the overall agenda, any additional information and registration link. These are the very basic things we would prepare. After
that, these are the process I typically have to do and I have turned it into an AI workflow. First I would need to brainstorm event titles to come up with the name 10X podcast and webinar content. With AI like this webinar, I actually have the assistance from AI to help me brainstorm these names. And I can pick the best one to move forward with this step. Human input. And then it will help me to write the event description also based on my own brand voice and template. And then it helped me to generate a logistic
prep checklist. So this is actually very useful because sometimes we just forget. I forgot some very basic things to prepare before a webinar. Things like preparing the slides or setting up the live stream on YouTube or schedule a zoom meeting, things like that. So I just want to have a very basic and straightforward checklist generator like this to help me come up with what I need to prepare for each webinar. And then I have this. The rest of the steps would focus on promotional content. So I typically have like three kinds of posts I often
need to do. One is for the brand page, so this is the LinkedIn profile of MindPal. And then one is for the community. So we have our own community of MindPal users and also a few other external communities. So the kind of style I write to this community would differ from how I write for a brand page. So that's why I divide them into two different steps. And also I have a social media post for speakers personal page too. So for me I also have my own following on LinkedIn. So I often create posts, but
also in my own style. So to post on my my post. And I also saw this useful for many of our customers who have guest speakers because they can prepare like a media package or something like that to give to their guest speaker. So the guest speaker don't have to prepare a lot, they just need to take the post here and paste on like post on their own social media channels. So that's very convenient. And I have this invitation email too to send to some of our past event attendance. And this step is not applicable
to MindPal, our team specifically. But I know that many of you guys often have to prepare event registration page, so you guys have your own landing page. And so this is something I have seen people doing a lot too. So we can generate the copy of the landing page as well and then an email sequence for the participants, including reminder, follow up, things like that. And I also seen many of my customers produce an email sequence for the guest speaker too. So besides an email sequence for the participants, you can add email sequence for guest
speakers because you would often need to remind them or maybe give them a guideline about what to expect, expects or maybe give a thank you email to them afterwards. So things like that. So this is the AI workflow that has been really helpful to me to prepare for these webinars. And this is an example, this is an example of how I use this webinar, this workflow to prepare for this webinar. So I provide the information, the brief here, which is AI for podcast and webinar content, all of the basic stuff here. And then. So this
is based on my discussion with my team. And then after having all of this in place, this step, brainstorm event titles for me. And since I am quite satisfied with this event titled Brainstorm Results, because I have trained it about how to choose event titles, I need it to be very straightforward, descriptive, don't use both words and all sorts of things. I don't want it to include any ambiguous words like transform or something like that. I really hate those kinds of words because it sounds really cliche and cringy. But thanks to that, I'm able to
generate these very quality event titles. And I just picked the first one because I really like it. It's short and straight to the point. I choose this one as the final title, as you can see. Then it helped me generate the event description based on all of the information above. This is by Cloud 3.5. My favorite model for writing this is logistic prep checklist. Some very basic one like live stream setup on YouTube and Zoom, stuff like that. And this just helped me to make sure I'm preparing everything I need because I also need to
do a lot of other tasks and there's no way I can remember everything without a checklist like this. So after the checklist it also helped me to write the social media post for brand page. This is the brand voice of MindPal. I really like it. Concise, straightforward, this is for the community. It would be like a little bit more conversational and casual. Like as you can see the language here, you will walk away with something like that. So when I post into a community, I also have the tendency of using images as separators of content.
Because when I post on YouTube, on Facebook, I often found that if I don't have these images at the first as a first sentence in a section, it's really hard to navigate because yeah, so that's why I taught it to make sure to include these images. And then this one is for the speaker's personal page. I use it to post on my own LinkedIn. And then this is the event invitation email. And then this is the event registration page. You can feel free to teach it in to generate the copy in your own style, in
your own structure, the structure that you prefer. And Then this one is for the email sequence for participants pre event reminder email post event, thank you email follow up resources email. And yeah, so in. So typically I would need to spend maybe hours on this process, but now all of all of this takes me just like two minutes to produce. And this is very awesome because I just spend something like 30 minutes to set this up at first to make sure these agents write and work in the way I want. But after that I can
repeat this for any webinar that I need to host. So it's very convenient for me. So yeah. So duong, what's your comment about this kind of workflow? What do you think how can we further improve this or is there any particular thing in this workflow that you want to highlight for the people in this webinar about how they can implement this in their business? I think the the most important thing to for example someone duplicates this template workflow from you and then use in their own business is that we we will need to using our
own brand voice in every in I think most of the step here. For example you show that we have the content for the brand and content for the personal profiles and content for the community. And for example, if the host from the podcast is me, I will run a kind of post that look like me writing it and I need to assign the different brand voice to the agent in that step to try to mimic my startup writing. And I think even for some things like the description and for the copywriting for the event based
because I think every brand will have different way of structuring content for the audience. And I think that is the most important thing here. Maybe to further improve this workflow. I think maybe the only thing depend on the business we can customize that for the email sequence. Maybe some of the business will have a more sophisticated email sequence. For example the email sequence will show have two emails. So I think I have three emails. The invitation, the reminder and the recap and thank you. But maybe some business will have maybe even 5 reminder or 3
reminder before the webinar for an important webinar and then we can break down that into some smaller step into that or even for some serious content procedures webinar. We can have some kind of like advertising and we generate ad copy for Facebook ad or Google Ad or linking app. We have an ad campaign generator workflow and we can copy some of the step in that workflow into this workflow. I think that is some of the idea to to like further like fine tune this workflow to Fit with the purpose of ending a different business. Yeah,
and thank you Dong for the insights. And just to show it to some new some beginners here, Wen Tong mentioned brand voice. Don't refer to these settings, the brand voice settings. In output settings of the agent, you can create any style that you want. You can just paste some existing content that you have already written for each format and the AI will automatically learn that style and it will try to mimic that style if you connect the brand voice to the agent here. As you can see, we have the LinkedIn style. So this is trained
on some of our best performing LinkedIn posts. So we use it all the time. And I just want to quickly highlight one another thing here, that which is the human in the loop nature of this AI workflow. So as you can see for this AI workflow we start off with provide some information and then the AI brainstorm the event titles for you. But by adding this little human input step here, the AI workflow will automatically pause after this first step to ask for the selected event title selected by you before it moves on on the
rest of the steps. I think this is a very like this is a very simple yet powerful practice for MindPal because just last week I got a question from one of our customers about now he can force the AI workflow to pause at certain steps before continuing with the rest of the steps. Because there are some steps he prefers the AI to just go on and continue doing all of that without his oversight. But there are certain steps like this one brainstorming. He would need to step in and decide which is the final item to
go with. And with our human in the loop feature on my pal, you can very easily do that by just adding a human input step anywhere in the workflow. That's it for our first AI workflow end to end event planner. The next AI workflow that we will look at is this podcast guest researcher. So after we have prepared and planned for the event, the basic information I know many of our customers podcasts and webinar content involve an external guest speaker. So they would invite guest speakers to the podcast or the webinar to discuss a certain
topic. So one thing they need to do all the time is to research about the guests and compile the insights, select topics and formulate the questions as you can see here. So what this AI workflow does is that you can provide the link to this guest profile. It could be their LinkedIn profile, it could be their own website or something like that, and you provide the episode theme as well as the target. And like after all of this, you can save a lot of time with this process. And I One interesting thing that we just,
we have just recently introduced in MindPal is that you can actually connect Perplexity to this process. So I know many of you are Perplexity is your favorite way of researching. So we have recently introduced a beta feature which is to connect Perplexity to your AI workflow. So Perplexity as the AI model responsible for the research steps. So we will share a detailed guide on how to connect that into your account later on. But it's very straightforward and this is when it gets really powerful. It means you can get the great research capabilities from Perplexity for
research steps, but you can involve all of your best performing agents for the other steps like question formulation and all. So this is an example of how this AI workflow takes place. So this is since we don't have any guest speakers for our webinars, I just take an example profile here and this workflow would research the introduction. And as you can see here, it's actually quite comprehensive. It takes this website, but it also identifies a couple of other links here. So from now her LinkedIn profile and stuff like that and it go through all of
that to prepare the to research about this person and then it compiles some insights that are relevant to this person and select suggest some topics that are that could be explored in the podcast episode as well as formulate the questions here. So this, this is a very straightforward yet I think this is something that could be very useful for people whose podcast or webinar series involves working with guest speakers a lot. And yeah, I think tong, I know that you used to help one of our customers to set up a similar system for his podcast
involving interviewing book authors. Right. Can you share a little bit about your experience and how you think businesses can apply this AI workflow into their business? So in the book Auto Interviewer, the the way he doing research is is quite similar, although the specific detail is a little bit different. Instead of just the profiling of the other guest speaker, we will involve multiple books from that speaker into the trigger input list in the workflow. And the agent will need to decide it reads the profile. You need to read the book too to extract some of
the inside in the book that will be prepared for the question list and the topic brainstorm. And when it can do that, it will show to the speaker that he read the book because we can take the exact reference from the book into the question. And so the Answer will be more insightful because that is the like the children of the author. And there are several more steps to better help us to do research. For example, when we brainstorm the topic of the. Of the podcast, we actually can have a human in the loop step
after that. So we can select, we can select the topic that we want to. We want to dive deeper on. And after we get about the topic, actually we can using those output as the input for the previous workflow to prepare for the some of the copywriting and material before we doing the podcast and some of the material we will send to the guest speaker. I think that will be the way to. The way to best customize this workflow for the business. I think the principle here is that this workflow is logged into read everything
related to the guest speaker on the Internet or any kind of materials that we can find and then summarize and compile into a list of actionable step or some actionable insights that we can use as the material for us to prepare for the podcast. Yeah, thank you so much for your comment. And I also used to support some of our other customers building their own research system for guest speakers like this. And I think the. One of my key lessons is that people would need to really understand their. What they need how they often do
this in order to translate that into an AI workflow that actually works for them. Because the way people do these podcasts and webinars can be vastly different. For example, one of the customers I support, they also interview guest speakers, but they focus on business owners. So they always have a step aside from researching about that guest speaker, that person. They would need to also research about the website of their company to understand the background of the company as well, because it's a podcast about business owners. And then for that, that podcast also have a very
specific format. So they typically would start with some interesting facts from this person. So he always has a step to look for some interesting angles to start the podcast with. And that's very helpful for him to pick the best angle to kickstart the podcast. And then they would go on with the questions. But at the end they often have, I think he calls it fire burning questions or something like that. Essentially, it's a series of quick questions to ask the guest speaker before they end the podcast. So he also has a step dedicated to coming
up with that burning question list. So yeah, I think the key thing here would be you can feel free to customize this AI workflow to fit how you, how you typically do these Podcast episodes and webinar events with guest speakers to make sure that you get the help results as you understand your business the best. So feel free to customize it so that it works like it fits better into your use case. Okay, so it's time to continue with our next AI workflow here. This is also for free event. This is a webinar content prep
I often use before each webinar. So essentially I provided all the topic, target audience agenda and any raw thoughts I have about the content. Like some raw plan key points I want to discuss and I often for every webinar I often ask users about what kind what they would like to see ask or during the webinar. So I also compile all of the questions here so that the AI would be able to analyze all of this and suggest an outline for the webinar write some rough, rough scripts especially for transitions part because I realized that
I'm not very good at transitions so I often take more time than necessary to transition between sections. So that's why I prepare an agent to help me with just thinking like going through the whole agenda and coming up with all of these transition to make sure I do it well. And then it can also identify some interesting angles to explore about this topic and also prepare some frequently asked questions that we can expect to be asked by the audience. So we can best prepare for these questions as well. So that's how like so this, this
is how I prepare for today's webinar. I also quickly go through all of this and it I provided all of the information together with some raw notes of me and as well as some questions from the audience. And then this AI will compile all of the insights and analyze all of them to identify the key themes here and then it suggests an outline for me. Of course this is up for me to decide whether I would like to go forward with this, but I think this is pretty helpful for me to identify some key talking
points during the webinar and then it helped me to write a rough script for the transitions between different sections and then it identifies some interesting angles to explore. So I also identify some interesting angles here like about quality control, how to maintain brand voice across AI generated content I think which is one of the main concerns of people when it comes to content marketing like this or it's about time roi how to convert one hour recording into weeks of worth of content. This is also a very interesting, interesting angle to present and I think this
is something that can resonate with people as well. So yeah So I think this AI workflow is pretty good at helping me have a clearer idea about what to expect in the webinar. This is some frequently asked questions that the web, the AI workflow thing that you, that I can prepare based on the topic. I think this makes a lot of sense. For example, how to use MindPal's webhooks for data extraction, how to get the content from somewhere else maybe, or send the content to somewhere else afterwards, how to customize the content for specific audiences
and stuff like that. So this is what I think this is a very somewhat personal AI workflow. But I think it is very helpful for hosts or speakers to prepare. And it would be even better when we teach these AI agents the business context about your business. So that those suggestions here would be very unique and specific to you. It would provide suggestions that actually make sense. I would recommend you guys to take a look at this use case too. Not just for webinar, but also for all other things that you do. And I think
AI is such a great way for brainstorming and just having many perspectives into a topic. And it's like having an assistant right beside me, just like a consultant. So I really like these kinds of use cases. And I think we can go to the next AI workflow, which is the one that I'm pretty excited to share. This is the webinar end to end webinar recap generator. This is, I'm sure this is something you guys do a lot. And for this AI workflow, this is a massive AI workflow we have right here. What this does is
that we provide a webinar YouTube link. So typically I often upload all. I often live stream these webinars on YouTube. So I always have have a YouTube link. But in your case you may have the recording file instead. So feel free to change the 5 star file 5 type here and then it will do the full transcript for me. And then first of all, one thing I often do is generate the YouTube chapters. This is very daunting if I do it manually. So I, I use it to. Because the transcript we get from YouTube always
have timestamps so we can use it to turn into YouTube chapters for me. And then it generates those call to action comment for the first comment. So I often put the first comment. So this would identify the most important call to action based on the transcript and prepare that comment so I can just copy and paste. Very handy. And then it would extract some memorable codes for posts from this webinar. So based on that one, our recording. Chances are we have a lot of codes that are important and easily could be shared easily. So this
one would help extract some memorable quotes from that. Another thing is that I often extract sections of the recording into separate YouTube videos. So for example, in our last webinar about how to build AI internal AI tools for your business, I break down the big recording into one video for live demo of MindPal, one video to explain what is the difference between MindPal multi agent workflows and cross app integration platform like Zapier or Make and things like that. I often have this because I have two different channels, one is MindPal and one is my own
YouTube channel. So I just, I want to maximize the content. So I have this step. Another step is extract snippets for short clips. So this is something I saw a lot of people do. So they have this big webinar recording and then they extract the snippets that can be turned into a TikTok video. And oftentimes they have to put it somewhere and send it to their video editor or VA to do that. But for this, this stack will automatically look through all of the all of the transcript and it will extract the snippets that are
worthy of TikTok viral video. And by worthy I mean it needs to be like it has good content and it needs to be understandable even on its own without the phone context of the webinar. So yeah, so this is the step that I see people do all the time too. Another thing is writing the caption for these short clips and then we have these steps to write the recap post. We have the recap social media post, but we also have the recap email and we have the blog post as well as the ebook. So the
ebook is something, to be honest. This is a new, this is marketing content format that we haven't explored at MindPal a lot actually. But I really want to experiment it with this. So I think one hour of webinar recording like this webinar we are doing right here, it contains a lot of value and all of this value can be turned into an ebook for people that prefer to read the content instead of watch the recording with this AI workflow, this is how I generate the recap content for our last video, our last livestream about how
to view AI internal tools for your business. This is the phone transcript here very long do all of this here. This is more than one hour long webinar. So that's why the transcript is this long. But as you can see, it's very long but it helped me transcribe phone so that's great. And then it generates the timestamps for me. Very good too. It got the name wrong but it's okay. I can change that very quickly or I can just put it in the instruction the brand names is MindPal not MyPow or Mpow like this but
that's okay. I think the most important thing is that it got the chapters right and it can suggest me the timestamps too. So great. And then this is a call to action command. This is some memorable codes taken from the transcript itself. I really like it. And then this is the sections for separate YouTube videos. So understanding my pal all of those things. I think it also sounds good. I would pick some of this to become the actual YouTube videos. This is study for snippets for short clips. It extract the snippet here with timestamps and
then this is a transcript. This is cloud 3.5.net and again I think it's excellent at understanding the content and pick the most worthy content for TikTok videos like why multi agent AI is better than single agents game changing canvas feature how my power differs from custom GPT real business success story, the power of brand voice, stuff like that. And I think each of these is actually yeah they are great topics and I think I really like this or AI models working together. A GBT model and Cloud3 profiles are networking together. Yeah, that's great. I think
it's excellent as suggesting these snippets for me. And this is the caption that will go along with the short clips here. And then this is the recap social media post. This is the recap email that we will send to our audience or put it in a weekly digest email of MindPal. This is the ebook. So for the ebook part I think the content is okay. I'm not very satisfied with it yet but I but this is a work in progress. I would need to continuously improve on this but it gets the key content correct. Now
what I need to do is to improve on the writing style and the. And the way it present information so that it be that it could be more descriptive and informative in my experience in my perspective. But I think it get all of the information correct here and that's. That's a great starting point. And this is the blog post based on the content of the webinar. Okay. And like as you can see this is a very. A lot of things here and. And I think this, this can. This work process alone can save you so
many hours. And if you can just spend some time train an AI workflow to understand how you work and what your how what kind of output you expect, then then you the the reward you can get is that you will get all of this in just a matter of minutes, not hours or like before. So I think this is definitely must have AI workflow for all of the podcasters and webinar hosts out there. So yeah. So Dong, would you like to add any comment in this AI workflow? Well, I think I have no comment. I
think this is so comprehensive enough for any comment here. I think you can move on to the next workflow. Okay. And all of these AI workflows are available as templates after the webinar, so you can easily duplicate them into your own workspace for further customization and training. So make sure to stay tuned. And we will post all of the links on our YouTube as well as in our community and in our weekly digest email this Friday as well. Now this is the next AI workflow that I use. This is for webinar performance reviewer. So essentially
it takes the webinar recording and any participant comments. So you could do a survey or something like that to collect people's feedback. Now you can just upload all of them here and this AI workflow will evaluate everything from the content, the delivery, audience feedback analysis and suggest actionable items for you to improve for future webinars. So I also use it for my last webinar and this is the transcript, this is the evaluation. So it essentially score, it gives me a score across these different aspects and then it identify pinpoint the weaknesses for me like which
is when I where I recognize that I have I'm not not very good at transitioning. So some sometime I pause and it took more time than necessary for the transitioning part. So I also improve that and this is also where I learned that it would be great if I can have some sort of recap key highlights after each section so that the audience can stay focused and easily have a big picture of things more easily. This is the audience feedback analysis here to identify if there's any positive negative feedback and things like that. This is
the overall impression key takeaways, the areas for improvement. These are some action items for future so that we can look at and it's divided into three categories, high impact and feasibility, medium impact and feasibility and lower impact and feasibility. So I can always take a look at this and know how I'm doing and then I will improve for my future webinars. So I Hope that you guys really find that my webinar hosting skill has improved over the past three webinars. I really hope that. But I know I will continuously improve my skills with the help
of this AI workflow and yeah, okay. And we are finally arrived here as the last AI. This is not an AI workflow. This is just an AI chatbot actually. And this is something I have learned from some of my customers who have, who often have a big library of digital materials. So they can have webinar recordings, ebooks, books, guidelines, a lot of things. And what they did is that something I found really interesting. So they have like a chatbot that have been trained on all of these materials and it can answer the questions, but at
the same time it can direct people to watch this episode or look at this ebook specifically. So for example, this, this is. I asked this question and it referred me to this part of the YouTube channel. So yeah, this is app. Let's wait a little bit. Yeah, I'm not sure if you can notice that, but this is exactly the part where I discussed about how Zapier and Mate connect apps. But MindPal connect AI agents and we are more optimized for AI driven tasks like content generation, evaluation, analysis, research. So yeah, so I think that's, that's
been. That's quite helpful. And then if you. This is like we all only have one previous webinar recording here, but imagine if you can just input all of the webinars you have ever done, all of the news newsletter issues you have done, all of the maybe podcasts or ebooks you have done with the links, of course. And this AI chatbot would be able to answer the question, but it will also direct the person to. If you want more information, you can watch this webinar recording or watch this podcast episode or take a look at this
ebook. So it could be a great navigator for your audience and it's a great way for you to engage with the audience soon. So I think definitely something you can consider if you have a lot of digital materials. Yeah. Okay. I think we have covered all of the six use cases related to podcast and webinar. This is a lot. So we are almost near the end of the webinar, but we will stay around a little bit for some, for some questions from the audience. So Dong, is there any question from the audience we can answer?
Yeah, so there are some, some question that we already have in the. In the form and some question here in the chat. So I will ask you one by one. So in the Form NIDA have one question about how to fine tune what is presented to the preschool education teacher and mother. I think so the question here that how we can customize the output of chatbot or workflow to a specific audience and in this case it's a preschool education teaching. I see. Yeah. Thanks for your question. So to improve the output to tailor it to
your audience, there are two main things you can do. One is there's a design output format section. That's where you can teach it the format of the answer that you want. For example, if you want it to be concise within 100 words or something like that, you can specify the format you want there. And then in the output settings you can train it on a specific brand voice writing style. So it could be. So it's. You could provide it some sample emails that you written to your audience before so it would notice the kind of
the language and the tone you use to communicate with your audience. And it will try to imitate that during its interactions with these with your audience. So we have a video on how to humanize AI writers and it dives deeper into the topic of how to teach an AI agent to write in the specific style you want. So you should definitely check that out. Yeah, you can show human AI content or, or this AI agent who follows our writing style in in our YouTube channel and you will find several videos talking about that topic. And
we have another question about the most effective way to generate podcast contents. I think that is a very, very big question. I think is depend on on the kind of audience you are approaching. But as tom show in all the workflow, this workflow can help you to prepare and research about different insights, angle and information and other stuff you need to best prepare for the podcast and the webinar. Yeah, actually I think this question also raises an interesting topic of AI generated podcast content. Like what we have seen with Notebook LM for example when you
can generate the conversational style conversations based on a content you have. Actually this is not the topic that we discuss in today's webinar. Today's webinar will focus is more on how to enhance your existing workflows. So the actual webinar and podcast is still delivered by human. But how you can optimize the workflows before and after the podcast so that you can save your time and focuses on the like the actual thing that matters here which is the webinar or the podcast episode itself. So but I think you can share, Dongyu can share a little bit
of your Thoughts about how this is heading like with the advancements in like I know Notebook LM and then some days ago 11 labs also just released a conversational API for something similar to this. So what's your thoughts in this? I actually don't realize this. I, I think I I I'm not referral to like social media or podcasts like Spotify or YouTube to listen something AI generated and I, I just shoot up because for example sometimes I not the podcast but sometimes I, I, I read some comment from people on for example Twitter or LinkedIn
and after a while I realized I feel so dumb that I read a ton of AI generated content from that. So I think maybe that is a way to produce content on some specific purpose but actually to consume the kind of authentic content. I I, I, I don't feel that the AI generator podcast is a thing although I think that it will blooming in the next I think not even future even few months because now the barrier to to entry to the podcast will be easier and even in my town for example in your workflow
if you connect another agent after the transcript preparation agent got connected to the OpenAI take to speed, you can actually generate the podcast from jot and script in the workflow and you will have something prepared to publish. The quality maybe some questionable here but actually you have a thing to share and but I think I don't like it. I think that is not the way we should heading for the content and maybe I think some of the platform will have some kind of regulation about those content. I see. Yeah, this is an interesting topic, right?
I think it's very interesting to see where this would be heading in the next few months. Yeah. Any other questions? Dong? I see several question in the form here but there are some question about how we can use webhook in my panel to send the data, the output data of the workflow to other automation platform and how we can create a workflow. I think I can quickly answer this question because we have other questions in Zoom and in YouTube console. So for Webhook we have a very detailed tutorial on how you can use webhook and
another very detailed tutorial on how you can use webhook with Mike and we are preparing to make another video on how to use web with Zapier. So when you go to our YouTube channel when you can search for example make.com or as you can search for webhook you will see a bunch of tutorial about that topic and it's going to be more detailed because it is I think a little bit more technical to be discussed now. And the question about the workflow creation process are also available on our YouTube tool. You can basically we have
three ways to create a workflow. The first way is to just tell AI what you want to automate and the system in my part will generate the workflow for your a baseline for you and you can further edit the workflow if necessary. The second way is by choosing a template. On our template workflow we have a lot of template in different category in business operation, in SEO, in podcast, in marketing, in education and other things. We have a lot of options for you to choose from over 100 templates and covering a lot of topics that
could be automated by AI. And we constantly update the list of options in our template library. Or you can upload an existing workflow because in MindPal people can export the workflow to be too to a file and you can upload that file again into the account. So the workflow in other people account will be lost when you upload the file. That is basically three ways you can build a workflow. And when you search Workflow 101 in our YouTube channel, you will see a quick tutorial, very detailed, covering different configuration and setting of the workflow so
you can fully customize for the use case. Yeah, yeah. There's a question from Nada now about can we put the keywords for the videos we recommend? I think the best way is that we can put all the links to the related videos on our YouTube live stream recording. We will include all of the related videos links together with the links to these AI workflow templates. So you can just click and duplicate them into your own workspace. So that's the best way. Okay, yeah, we are almost 10 minutes over the time, so thank you so much
everyone for sticking around. Is there any final question we should answer or can we. Yeah, we, we. We still have some, some question here. Okay, quickly. Yeah, I. I will select some question here. So there's a question from Kel. Let me summarize because the question is long. Okay, so I think the question from Kelly is that how we can structure our prom both in the Asian micro and the design output format in and the prompt in the workflow step in the workflow because she observed that we often have some kind of special syntax for writing
the prompt. For example, some of the syntax will highlight even this information or have some categories like rows, tasks or context. And she want to know how we can best use this special syntax injection in the prompting in my car. In general. I think you are the best at answering this. Yeah. So actually there is no concrete and specific formula for you to write a prompt and some kind of syntax that we divide into categories like background or task or context. Or we highlighting some of the important information or we write the importance often time
that is some of the things that we want the agent to focus on. And we have and we will have some kind of special syntax to highlight. And the way we are highlighting is because the prompting practice that is just the convention the best practice from other people, not the ways necessary to follow that. But the principle here that I observed after many times working with AI is that the important is that in the prom there will be some part you want to we want the agent to focus on. And in those parts you should
have some special syntax or wording that let the agent know that just what is important. For example, if I build an agent to write. To write a LinkedIn hook for me that I require it to be less than 10 words, for example, I don't want it to be more than 10 words. And the. And the important thing is that the output have not too long. And I will put the syntax. For example, I will have the syntax important and I will write in uppercase important and I will write the instruction maximum 10 words for example
and session should follow that. So that is the way you do. And even when you see in our in. In our prompting of the AI generated workflow or AI generation we we have often have the the. For example, if a step in the workflow referring the output of the another step in the previous step, we will often have a syntax given the following information. And in the. In the belows we often have like the title of the workflow step previously and we mentioned the variable of that value. So actually you can always feel free to
remove that part because that is just the convention when we writing the prompt to make it easier to read and convention with the agent. Because as we will have even more powerful last language model in the future, prompting actually will become easier. And the principle here is just very simple. You should prompt the agent to do just one thing. If possible clarify with the agent what is a step by step guide to do just one thing. And if are there any important or are there anything you want the vision to focus on? You should have
some kind of special syntax. So the agent will notice some difference in that part and it will try to follow the important note in the prompt. Yeah, thanks for the very detailed answer. Any other questions we should answer I think we have answered all of the questions in the Zoom chat so I think we are good. Yeah I think we only have the final question that is from a new customer from Venezuela, Eduardo. He asked a question about I think not really a question but he said trying to figure out what are the best workflow
for consultancy agency. Yeah. So to be honest, when I encounter these questions, as much as I want to help, I would still need more contacts from you in order to give you the best recommendations. I think the most important thing when it comes to implementing these AI agents and multi agent workflows into your business is that you would need to be the one to codify your business processes first before the AI can understand and do it for you. Some so my recommended so if you want some inspiration and some ideas about how you can use
MindPal for your business, we got a vast library of templates you can take a look at. Kelly just had a very great suggestion. Our Facebook group is also awesome for discussions like this where you can post your business and people can share their insights of how they use MindPal in their own business which is really helpful. I think as soon as you have understood about what processes you often do that you want to automate then building these systems on MIPAO should be very straightforward because as we have shared we even have an AI generator for
AI agents and multi agent workflows. So if you know your process really well it should be easy to implement that. Yeah, yeah, yeah, yeah. I I think I I have a very generic answer but as I observed many of our customers agency tools and Tom says you need to know the process first what you actually are doing. Normally you are doing what type things and then you transform those process from the work into the AI workflow to do that for you. Typically for agency I often see people have been implementing three kind of workflow. The
first one is the workflow for kind of lead magnet workflow to attract new customer view something that is helpful in zones. For example just I think two or three days ago I see a customer building a workflow as a lead magnet to help people identify their ideal customer profile in many different angles. That workflow have 70 step very detailed and very comprehensive workflow but is driving a lot of lead for his business and in that way he can necessarily showing the expert in the field and I think in the target audience. Are there anything that
could be helpful for them that could be. That could be translate into a workflow? You should build that and give it for free for people so they can start to know about you. The second type workflow is that some of the workflow for internal process, for example writing content for SEO for repurposing stuff for research. Every type thing that you think that could be codified into a workflow should be done here. And the final thing is that the workflow that view for other people view for the actual client, not just for lead magnet purpose. And
in that option we we have the payment support in the workflow for you to easily test your client. And even just a sneak peek we we are preparing to launch some support in our team workspace that you can add the client as a guest in the team workspace and you can charge them as a subscription or something like that in the team workspace workspace. So for example you have an agency you can create a team workspace on MindPal follow agency. You have some core member and then you have some client going through that and you
can charge them a fee, a subscription or any kind of fee that you deal with them and onboarding them using those workflow that they get access to in the workspace. And I think that is the overview on what we often see people are doing and people are planning to do Azure Agency on my channel. Okay, thanks for your answer. I think we are. We are. We have answered all the questions we received today. Thank you so much everyone for sticking around and we hope you have found today's webinar helpful. The recording is available on our
YouTube channel in the live section for replay and we will include all the links to the related videos we mentioned today together with the links to all of these AI workflows in the video description and in the first comment of that live stream too. So thank you so much everyone for joining and we will see you in our next webinars. Take care. Goodbye. See you.
Copyright © 2025. Made with ♥ in London by YTScribe.com