Get Ready With Me: Generative AI Webinar Prep, Part 4 of 5

In today's episode, this is part four of my get ready with me series on how I prepare for a webinar using generative AI. So let's continue. One of the things that is challenging about the consumer generative AI web interfaces is they often can't handle large amounts of data in a prompt itself. The developer versions of software can are will do typically a much better job of this. So we see here now it's extracting out the key themes from this data set.

So we're going to go ahead and take a screenshot of this. And now let's duplicate this slide and change the opacity of this so that this becomes more prominent. All right, another example. Hotels have property management software. Let's see if we can find some.

What is hotel property management software? Let's take a look at some images. Ah, look, we have have a what we got here. This is a an image of property management front desk operations interface from Crossdesk. And this is room.

This is room stays. So let's open this image in this new tab. For any kind of extraction, we'd want to work with a model. We'd want to work with something where we can extract useful information from visual identification information. So let's go ahead and save this.

Let's save our now to create a new prompt. I have attached an image of the front desk operations for our hotel from our PMS or property management software. The image shows our capacity. Which guests are booked for which rooms. Provide an analysis of our property based on the image.

How are we doing in terms of keeping our rooms full? And now we're going to take that image that we saved, hit run, calendar view style room occupancy, dates, high occupancy, possible trends. Several rooms are uh locked with maintenance. So this is super helpful for um our example. Let's go ahead and clear the navigation out so that we can because again, we're not necessarily trying to say that we are experts necessarily in hotel and property management.

We do want to show our attendees, we've done our homework, and we can and we can give them examples of here's the software that you use in your industry, and here's how generative AI, even if it doesn't plug into that software, could be made to work with that software by taking things with screenshots and turn it into useful and actionable insights. Let's move on to our next section of the deck, which is going to be summarization. I like this, that's fine. This is summarize call customer history. Um, I'm gonna leave the hot the this in here.

I'm gonna swap this out. This is our example of uh monitoring uh YouTube channels, and we're actually going to should we do an example? We do a new example of how a hotel manager might be doing, yeah. We might want to. So let's do this.

Let's go to YouTube. Let's see and let's call look for hotel travel hacks. All right, and let me go ahead and close all these other tabs because we're gonna need a list of URLs so we're gonna hook uh pull out cheap hotel hacks hotel hacks for smart traveler all right so we've gotten a good selection of the major hacks we're gonna take that list of videos clean it up just a little bit here and now as we did earlier let's take out the closed captions okay so as we did before we now have our data file of the hotel hacks that people have come up with uh from YouTube so let's go ahead and get this in here today we're going to summarize a collection of transcripts about hotel hacks ways travelers can make the most of their hotel stays save money and do unexpected take advantage of unexpected loopholes outline the hacks discussed in this collection of transcripts in outline format ordered by priority in descending order priority is determined by the positive or negative impact actually this is the negative impact by the negative impact on a hotel properties revenues. Since our audience is obviously hotel folks, um, they would want to know what are people saying online? What tips are the audience being given that could negatively impact your business?

Okay, so let's see what's come up with for hotel hacks. Hacks that directly cost hotels money, booking cancellation, rebooking at lower rates, bidding for upgrades to first class, uh, using bereavement rates when you're not going to a funeral, that's kind of scummy. So I would definitely say that's that's worth that yeah people are awful. Um price matching with third party sites, finding owner direct rentals, student corporate rates, uh claiming corporates despite not being an employee of the company with a negotiated rate, booking mobile only discounts, using free amenities, etc. So yeah, this is useful for if you were a hotel property manager, this is the kind of thing that you would want to know what people are being told about what people are being told, what's in their media diet that could impact your hotel chain.

Remember we are if we go right back to our our key priorities list our key priorities list is uh recruiting issues, brand differentiation, employing experience economy technological innovation. You are using this falls into technological innovation and that very core hey we need to make more money remember in the the key priorities for hotels we need to make more money. Well this is how you're being cost money um and there you may or may not want to close some of those loopholes uh let's see I'm gonna get rid of this slide because we don't need it for this example let's go into our fourth category rewriting so NDA example is good the training and professional development's good, the that one's good. I want to do one for to address recruiting because I think that's important. So let's go to let's go to our our consumer model here.

Start ourselves a new chat and say, let's get our expert prompt rolling for what the Gen Z employee cares about. So we're essentially having the model export out its knowledge of Gen Z and the things that they care about within the hospitality industry. Right. So we have all of these um these key areas of the things that Gen Z would care about. Now let's go to D.com.

Let's go into hotel. Go away. And let's go into the Boston, Massachusetts marker market. And we have here the Elliott Hotel. Let's see if we can actually guest service agent Hilton Boston Park Plaza.

Great. Let's take this text. I'm gonna open up a text document here. Now rewriting from our preloaded prompt area. We're going to make a few little changes here.

Here's the job description from Hilton Boston Park plaza for a guest service agent. The job description is written for a generic employee. Examine the job description and then outline how well or poorly the job description addresses the specific needs and pain points of Gen Z employees. So that's going to be our prompt. Let's go ahead and feed this into Gemini and hope it doesn't choke.

And it says, hey, here's things that are good, emphasis on benefits, growth opportunities, diversity and inclusion, areas for improvement, lack of purpose and impact, flexibility, lack of flexibility, traditional tone, which is a big deal, um, vague career path. So let's say great, using the information provided and the facts in the existing job description, rewrite the job description to be more tonally appropriate for a Gen Z employee in the hospitality industry. Confine your writing to the facts in the original job description, because we don't want hallucinating things like benefits don't exist. That's not bad. That's that I would say that's an improvement.

Um if we had more information, um, we could certainly have this do a better job. And I would say that's the advice that I would give to any employer is to say, look, take the existing um job descriptions, take the known things that the audience cares about, and have these tools do a better job of writing your stuff, right? So if you look at this original text here, full job description. One of the most historic hotels in the city. So from old to bold, that's a nice rewrite.

Okay. Our fifth category, classification. So this is sentiment analysis. Hmm. We have those hotel reviews.

Let's do. And they are in JSON format, which is even better. Okay, so let's go and take our go back to Gemini. Save this. Okay, so we now have we've taken our JSON content set.

Here's 20 reviews, perform a sentiment analysis minus five minus five to plus five, and a large data set. And now we have a very nice sentiment analysis. So this will go into our slide deck as an example of how to classify text, which again will be super helpful for hotel managers to know. Our last category is question answering. Let's see, how do we want to tackle this?

Because there's a lot of really good examples in here, but these are all very, very real estate examples of how do you answer questions. I'm going to keep the agent in place. That is it for part four of my Get Ready With Me series. Stay tuned for the fifth and final part. See you on the next one.

If you enjoyed this video, please hit the like button. Subscribe to my channel if you haven't already. And if you want to know when new videos are available, hit the bell button to be notified as soon as new content is live.