Name: NotebookLM for the Win (Coffee & Clouds Series)
Uploaded: 2026-02-13T20:20:55.664Z
Duration: 22 min 45 s
Description: NotebookLM for the Win (Coffee & Clouds Series)

Transcript for "NotebookLM for the Win (Coffee & Clouds Series)": > Good day and welcome to our Coffee and Cloud series event here for February 2026. Our topic today is NotebookLM for the Win. I'll do a brief intro and we'll dive right in so you can get hopefully a first look or a new fresh look at the capabilities of NotebookLM either as a standalone app or as part of the Google for Workspace environment. We appreciate you taking time out of your busy day and week to be joining us here live or to be watching this on demand. A reminder that this webcast, as well as other events we do, are all part of our resource center, where we offer a wide variety of ways that you can read, watch and explore information and resources in more depth to help you make not only better AI decisions, but better IT decisions as well. So hop by the website at any point in time and visit there, please. If you have any questions during the course of the session today, there is a Q and A tab available on the right side of your screen. Click on that. That's the best way to ask a question. We can moderate and track, make sure we get them all answered. And a reminder that this session is with all of our live online events. Process the recording. We'll generally get it posted up on our website. Within about twenty four hours, we'll send you a link. Feel free to share if you've got folks who you think would find this information valuable. And so diving right in, let's take a look at why we're here. And the first question we want to answer is, what is NotebookLM? Basically, NotebookLM solves a great problem for small and mid sized organizations. When we hear about generative AI and AI, the big power of it comes from being able to build a learning model that you can use to make better decisions, analyze data, research, ask questions, get answers. And for large organizations, that's sort of always been the focus. And they've got the money to invest in cloud infrastructure and algorithms and data scientists and prompting engineers and data taxonomy engineers to make it all happen. The small and mid sized organizations, we don't have that expertise and we don't have the money to spend on that expertise. Notebook LM solves that problem because it lets you, either within Google Workspace or as a standalone app from Google, generate your own learning model. And you do that by identifying your documents and your content that you want to feed the model and train it, as well as link to external sources that you want to include. And what's great about this, particularly if you're using it integrated with Google workspace, it's secure and private. So it's with the same security infrastructure and permissions you have with your Google workspace domain. You have source grounding. You choose the sources that get included. Those can be dynamic sources that get reviewed and updated on a regular basis or static historical information you want in. But whether it's internal private data, which is kept secure and private, or you're accessing current information off the internet and the web in general, you ground the sources and decide what you want. It's tailored. I mean, you can create notebooks and learning models for different aspects of your business. You can have one that focuses on your services team or on your sales team or on your marketing efforts or on your financial performance or market competitiveness and competitive analysis. You pick the topic, you build the notebook LM, you use it for what you want to use it for. And it's integrated. It's integrated with all of the components of Workspace. So you can not only touch into your data and information and drive, shared drives and other components, but your results can be saved and published in many different formats. And, I'll introduce Bill in just a second as Bill gets into it. You can actually pick different different media for your output and how you analyze and present the information you learn. So at a first glance, as I said, it is a learning model development tool for you to use generative AI. Your screen divides into three sections, a left quarter, a middle half, and a right quarter. The left side is your sources, whether it's Google Workspace or from the web. The middle side is the chat. That's where you do your prompting, get your responses, do your refinement. And the right hand side is Studio, which is access to Gemini Studio from Google, which gives you multiple document and media output options for presenting your results. And I can go on and sort of talk about this, but I'd rather introduce you to Bill Seybolt. He is our senior cloud advisor. And he has demonstrated for us as an organization internally how powerful NotebookLM can be. And so I'm actually gonna bring him up on the screen and welcome, Bill. Thanks for joining us here today. Thanks, Allen. I appreciate that. And and you dive right in with the demo. This is what they really. came here for. Yeah. Absolutely. So as Allen said, you have, three different sections within note NotebookLM. And the way you get to NotebookLM is just simply by in your, address bar, just typing in notebooklm.google.com, and you would authenticate based on your Google Workspace credentials. And once you get there, you're going to see, like Allen said, the sources, chat, and studio when you begin to first start your your notebook. In this case, I have set up a a demo on the company called the company Crispy Crane, which is based out of North Carolina. And I have put in a number of different sources that already based on just things that I've been able to find within, whether it be YouTube videos, whether it be websites. And when I go to click add sources, here's where you're able to to to, like Allen said, be able to enter in dynamic sources or even a static static resources. So in this case, I can actually type in, for example, Krispy Kreme, and it's going to actually search the web once I hit enter and find the sites that have anything to do with Krispy Kreme that I can also then add to, add to my notebook. I can also come in and select, upload files if I have files that are local. I can go to specific websites that I know have that have addresses. And when I do that, I can paste in any links that are associated with these. And then on top of that, I can even go to places like, Google Drive and then be able to bring up, specific, specific documents that are available. So, for example, here, I have the latest, the latest annual report that's available that I'm now can make that a part of my notebook. Now underneath the the chat, I can begin to enter in specific specific queries just like I would if I was entering in something within, within within Gemini the Gemini app, proper. So in this case, if I wanted to just enter in how many doughnuts does Krispy Kreme make in a day, what it's going to do is it's going to based on all the information that I have, entered into my notebook, it's going to generate an answer based on that information. And it might take a little bit of time to pull that information up, but as it kind of processes that, the the one thing you wanna certainly be mindful of as you as you as you provide these prompts is that you can go as basic as this. And so here we have our answer right away that that typically, we're looking at about 72,000 doughnuts a single day from one location in Burbank. We have 5,000,000 doughnuts every day. It is it's kind of their their general guess. But in this case, what I can do is actually do something kinda along the lines of what Google recommends, which is is to to do a specific way of prompting, which is there's four aspects of that, is persona, task, context, and format. And these four are are kind of like a the who, the what, the why, and then the how of a prompt. So in let's pretend, in this case, I'm going to drop in, something a little bit more detailed around this. What I'm going to do is put in a a little bit more of a complex prompt that says act as a business operation analyst, and I want you to estimate the total number of doughnuts a single Krispy Kreme factory store produces. So I have a little bit more information here based on those four key aspects of a prompt. And if I click on, enter here, it's going to then begin to pull together all of that information and also put it in a way that that I've called it to add to generate within within the within this prompt. Now it might take a little bit longer based on because I've added a little bit more information around that, but you're starting to get a little bit more of what you're exactly looking for as opposed to just simply entering in something very generic in this case. So as you can see here, that it actually highlights, for example, the differences between a premise an on premise retail store versus a wholesale distribute distribution store. And it takes into account hourly production rates and even peak production capacity. For example, it can generate anywhere from 3,000 to 5,280 donuts per hour. Now another thing that to keep in mind is is that it's it's going to act it's going to, drop in information that is specific to what you've already fed it. So in this case, if I just put in a profit and said how many doughnuts, does a Dunkin' Donuts make in a day, What it's going to do is it's going to look at all the information based on Krispy Kreme that I've added in. And if I it does have that information, it will provide that for me. But in this case, because there's really limited information around that, you're only gonna see what it says around Krispy Kreme. Another part of this is that is that you can you can generate, information just based on for example, in this case, I'm going to say, hey. Give me a little bit of a competitive analysis between, Krispy Kreme and Dunkin' Donuts based on the information that you have, which is a little bit better of a prompt that I would want to have given that it's it has some information that is based on the both the comparison between these two corporate solutions. And in this case, you're gonna see a the what information is generated here is is based on the particularly around what type of stores there are. So for example, Krispy Kreme operates in a hub and spoke manufacturing model versus the Dunkin' Donuts stores, which are typically a QSR franchise model. And and what I can do, what's nice about this is is that it will provide me the links to, with these footnotes based on the information that that I've entered in. And then on top of that, I can even export this out. I can more I can copy this and paste that into a Google Doc. I can also click save to note. And what that's going to do is that's going to actually take that over and save it into the studio. The studio actually has some really neat features that that the chat doesn't have. And and what I, a lot of these, I've already kind of went ahead and generated for our demo here simply because it takes some time to put that together. So when you are in the studio, you can create a podcast based on clicking on audio. You can create a mind map. You can create a video. You can also create an infographic or a slide deck, all based on the information that's available to you within, your learning language model. It's not just simply clicking on and generating a podcast or an infographic. For example, if I just clicked on an infographic and I generated that, I might see something along the lines of this. Now what I really like about this infographic is is that it it provides a tremendous amount of information around Krispy Kreme. And at the same time, it used the corporate colors that Krispy Kreme has in order to generate this infographic. I didn't prompt any of that whatsoever. But what if I wanted to do something a little bit different? What if I wanted to generate something that would, that would that would be aligned to our corporate colors within if we, as Cumulus Global, were presenting something to Krispy Kreme. Well, I could come over to this pencil icon. And as part of this generating this infographic, I could prompt it to say, hey. Use these specific corporate colors as you create this infographic. And just to show you what what happened in that case, I click on this this infographic, and here, you're gonna see all of the information that that I've asked it to generate based on doughnut production in Cumulus Global colors and not in Krispy Kreme colors. Like I said, another part of this is you can actually generate a podcast, and this is not just simply an AI voice that's reading through a a basic amount of information. It's two hosts talking about Krispy Kreme. Let's let's take a listen real quick. It's the experience versus the spreadsheet. We have a huge stack of sources today. Financial reports from late twenty twenty five, factory transcripts, logistics analyses, and, you know, the historical records of their previous crashes. Right. And they all seem to point to the same thing. Krispy Kreme isn't struggling because people stopped liking donuts. They're struggling because of physics. Physics, supply chains, and maybe, just maybe, a little bit of corporate hubris. May So one of the neat things when you do create one of these podcasts is that you can actually create this so that if there's an interactive mode to this, so that as you're listening to the podcast, you can ask it questions, and it will find those answers and reply back to you. Finally, you're able to, like I said, create slides as well as videos. Here's just a quick sample of what that looks like. Here's the video that, that that. this created. You know the name. Just hearing it probably makes you think of that incredible melt in your mouth original glazed doughnut hot off the line. It's a truly iconic brand, a global sensation built on a secret recipe. And one more thing here, we can look at the slide deck that it created. Oh, that's Excuse me. That's the blog post. And these are slides that are generated in PDF format that you're able to then immediately bring into Google Slides. Or if you have Adobe Acrobat, the whole version of Acrobat, you can actually make some edits here and there within the slide deck and then go from there. Allen? Great. Thank you, Bill. Really great demo really shows the power of sort of the three sections, building your own model, being able to ask the questions, do the analysis you want. You have the ability of going back and looking for information if it's not finding something of what you want and adding more in. And then being able to be creative and effective in how you take the results and the information and present it to others or export it and bring it in, bring it out of NotebookLM for other uses. And so this has been a great overview and a good sense of the power. Now, the beautiful thing about NotebookLM is basically if you've got Google Workspace for business standard or plus or enterprise standard or plus, this is built in. It's secure. Just like you have the standalone Gemini app, you have NotebookLM. It's accessed when we set it up properly and securely, lets you pull data from your workspace environment as well as move those results and all those different outputs there. And so if folks want to learn more, we've got a couple of resources I've put up on the screen. One is, our AI landing zone. It's a collection of a lot of most of the content we've created around AI and curated from other experts around AI. So it's a mix of blog posts, e books, white papers. We've done other web events, both Coffee and Clouds, as well as Third Tuesday at Three series events around this topic, which provides some great guidance and direction. And so the landing zone is there. A lot of the information in the landing zone was collated from our resource library, which is where we keep our e books and our white papers. We've got two in particular that directly relate to the topic today, Unleash the Power of Notebook LM. It's an introductory e book, as well as a e book called Google Workspace, a guide for using AI at work. This goes through each application within Google Workspace and explains the AI features that are available as well. One of the best resources we have, however, is our team. Bill is our senior cloud advisor, is expert at this stuff and can help you assess your use cases, your security infrastructure, and your education and training needs so you can get this up and running and part of your work processes. So at any time, feel free to contact us with any questions you might have or follow-up as well. The QR code on the screen or the Bitly address will actually prompt you to book an appointment with Bill or one of our cloud advisors. And we'll have that free, no obligation intro call. We'll discuss about your needs, your priorities, where you're at right now and what your next steps might be. And so on that note, we don't have any questions in the Q and A section. I'll give folks another minute or so if they have question before we go on. A question I do have a question for you, Bill. In your day to day, how have you been able to incorporate Notebook LM into your work as you have conversations with clients and potential clients about their information technology and services. So a couple of things. What I found really useful is, a, is being able to take information around a particular customer and be able to feed that within the notebook and be able to generate specific questions around that company and be able to find answers very quickly that I might necessarily have that that other learning language models necessarily would have right off the bat. If I'm looking at specific information, it's it's very nice to be able to dive down into that. Another part of it just kind of on the side is that I have been, looking through and studying for, various, certification exams with Google. And I've taken my training materials and just put it all into one notebook. And it's helped me to study and even generate, specific test questions that I can be that I can quiz myself on as well as flashcards to be able to to improve my learning around that material. Okay, great. Now we do have one question that came through in the chat window. We'll deal with that really quickly as we're to we're hanging to the point where we're running out of time. Someone has asked, can we say anything nice about the security of local data you share with the notebook platform? And so, by local data, we see two types of local data. One, it's the data that's already in Google Workspace. And the second local data would be something you might have on your computer or a local file store or file server that you are uploading into NotebookLM to train the cut your model. What I'll say nice about that is when you take local data that's secure on your server and hopefully properly permissioned and you move it in to notebook LM, the permissions for that notebook, you have control over. You can decide who has access to view the notebook and use it, who can modify it, and who can also administrate it and give permissions to others. And so it's the same security infrastructure that you use to secure data on MyDrive and SharedDrives. The same way you can manage permissions to your calendar, you can manage the permissions and access to notebooks and the data that you're storing within notebook OM. It stays private within your Google workspace environment. So it's private within your domain and it's only accessible to those that you share the notebook with. And you have as the user, you have control over that. So it's an important thing, and it's a it's a great question. Thanks for asking that. On that note, we're gonna wrap it up. We ran a couple minutes long, so I appreciate you folks sticking with us. These are meant to be quick, intense, and to the point. Bill, thank you so much for joining us. It's fun. This is, I think, the first time we've done a live demo in one of these sessions. I think we'll figure out other topics and definitely have you back to do more. So thanks for joining us. And to everyone, Thank you. thanks for being here. Thanks for watching live or watching this on demand. Have a great day, and we'll see you next time.