AICSS

Workshop: How to Use AI to Turn Data into Insights (AI + Customer Success Summit Session)

Bex Sekar
  -  
August 9, 2023
  -  
2 mins

Workshop: How to Use AI to Turn Data into Insights

Does your CS team struggle with translating data into insights? Are you trying to figure out how to get AI to do it for them (maybe even yourself)? Then you’ll want to attend this workshop, where we will practice using AI to create high-level takeaways and insights from data and its visualizations.

You’ll come away from this hands-on session with new data insights that you & your team can immediately use.

The below are not required to attend the workshop, but in order to get the most out of it:

- Bring a SQL query you’d like to optimize and work on

- Have a Google Slides or PowerPoint Presentation ready that includes data that you’d like to get insights on

- Bring a Google Sheet or Excel report of non-formatted, raw data you would want to leverage in a presentation

Speaker

Jenn Snider - Sr. Technical Account Manager at Matik

LinkedIn

Jenn Snider, a seasoned Sr. Technical Account Manager with an impressive 8-year track record. Having excelled at Yardi Systems and Atlassian in the past, she brings a wealth of expertise to her current role at Matik. With a passion for delivering top-notch customer experiences, Jenn continues to thrive in the ever-evolving tech landscape.

Watch Workshop: How to Use AI to Turn Data into Insights

‍Transcript

This transcript was created by AI. If you see any mistakes, please let us know at marketing@matik.io.

Matik MC: Awesome. Okay. Hi, everyone. Welcome to Matik, AI and customer success summit is the last session of what's been a really educational day. You are currently watching the workshop of how to use AI to turn data into insights. I'm very excited to introduce our speaker. And as I introduce her, I'm gonna have her answer our icebreaker question, which is what would you use AI to automate in your life outside of work? So we are lucky today to have Jenn Snider, senior technical account manager at Matik presenting.

Jenn Snider: Hello, everyone. Thank you so much for joining the session. I'm excited to share what we have with you today. And if I can use AI to automate a portion or something in my life that's a really good question box. I mean we're not there yet. On a funny note, blow dry, my hair would be amazing. But aside from that, I think I would have AI tell me… what grocery stores would carry sugar free bacon to sugar free and you can only find it in certain places, one whole foods, but the grocers near me, don't really have it. So if I could tell me the groceries within like a four mile radius, that would make my life easier.

Matik MC: I love that answer. But maybe instead of sugar free bacon, I might be looking for something else for me to, before I hand it off to Jen. Just a reminder for everyone watching. If you have any questions at any point in the session, go ahead and go to the Zoom control bar and use the Q and a feature. We'll be answering questions throughout the session as well as at the end. Okay. Jen, I'll hand that to you. Awesome. I will go ahead and share my screen.

Jenn Snider: All right. As I said earlier, thank you everyone for joining the session. We're excited for you to see what we have in store for you. List in front of you are the agenda items that we're going to be covering today, in addition to some time estimates of each exercise, so that you're aware of what's gonna come next and how long it's gonna take. Just as a quick reminder, make sure you are set up with the following that's an e-mail link to your Google account, a SQL query that you would like to refine if you don't have one that's okay, don't worry about that, a presentation you would like to gain insights from, and a data report either in a Google sheet or excel format. If you don't have any of these items, don't worry that's okay. We do have a sample presentation. I ready to go for you to utilize for today's exercise in addition to a sample data sheet report as well. You'll find a help documentation within the chat that has linked to everything that we're gonna utilize today, and also some reference as we walk through these exercises. Once we're done with everything, we'll go ahead and wrap up with a Q and a at the last 10 minutes of the workshop.

Jenn Snider: So at a high level, we'll be exploring some content today using the crawl walk run approach. So think of it as like taking small steps to harness the power of AI, we'll go ahead and start with the crawl phase which is something that you can get started with right now. And then we'll go ahead and transition into walk which is a bit more advanced but still very doable today. And then finally we'll reach the run stage where you'll see the full potential of AI in action. So let's go ahead and kick off with the crawl phase. And in order to complete this, we'll be utilizing ChatGPT to create and refine a query. So ChatGPT is an AI powered chat box that has gained significant attention for the sheer variety of comments and queries that it can respond to and provide inputs for. So, before I go ahead and switch gears and log into ChatGPT to start the first exercise, go ahead and log into your ChatGPT account. Now, if you don't have one today, you can create one, it is free. But once you log in, we're gonna go ahead and start with either option one or option two. So if you don't have a SQL query, we can, we're gonna leverage ChatGPT to create one for us. If you do have a SQL query and want to refine it and see what I do to refine a SQL query, you can do that as well. But for today's, workshop, I'm gonna go ahead and do both. So let me go ahead and switch my window. I'm gonna wait. I'll say like a minute just to make sure that everybody can log in to ChatGPT, so you're able to follow log. And while we're waiting for this one minute, if you have any questions or excitement that you want to share, go ahead and put it in the chat.

Jenn Snider: All right. So before I get started with leveraging ChatGPT to create a sequel query for myself, there's some things that I want to bring to your attention. And that's actually the limitations of ChatGPT. A couple of the limitations that I really want to highlight is the inability to ask clarifying questions. So if your input is ambiguous ChatGPT, my guess your intention rather than asking clarifying questions for a better understanding. So please be mica of that. Another thing I want to bring to your attention is the limited context window that CPT. The model itself has a fixed context window which means they may not remember information mentioned earlier, if you have a longer conversation which can lead to inconsistent responses. So please keep that in mind. With those two points brought up, I would like to highlight some best practices that will help you in leveraging ChatGPT. First be very specific, especially when you want a ChatGPT to create a query for you. And by specific, I mean list the table and call in names to avoid ambiguity, also clearly state any conditions, filters or sorting requirements. And lastly, the more precise you are in your description, the better the generated query will be okay. So with all those points in mind, let's go ahead and dig in and get started. Now, I have some prompts here that are already available just to speed things up because I am a slow typer, nobody wants to watch me type. So I'm gonna go ahead and paste in my first prompt and based call. I'm using a very easy example of customer orders. So what I mean by customer orders is I have an order table and a customer table within my order table, my column ID'S are my product and my price. And with my customer table, my column ID names are name and e-mail. I also let ChatGPT know that my customer ID and my order tables is my foreign key referencing the ID column and my customer table. So after I laid out all the tables and columns that I have, I'm asking ChatGPT to write a sequel query that's gonna find the total revenue of each customer by calculating the sum of their order prices. So now that I have that in the chat, I'll go ahead and send it.

Jenn Snider: And ChatGPT just created a simple query for me. So as we take a look at this, ChatGPT is utilizing the join and group by clause to make this happen. As we scroll down a little bit and take a look at the query and see what ChatGPT is doing. They're selecting the ID name and e-mail columns from my customer table, and also calculating the price utilizing the some function. It is also using a join to join the two tables because I referenced my foreign key and both my orders and customer table. And it's grouping by the results, the name and e-mail as you can see, it's saying here, just replace customer ID name and e-mail with the actual column names in your database. If they are named differently, you can then copy this code, put it in your database for or whatever tool that you're utilizing to search for those results, which is really cool. But let's say I wanna take this a step further. I have the information that I want but I wanna filter it down by customer location. The one caveat to that is I've already told ChatGPT my order and customer table which doesn't include location. So I'm gonna have to tell a CPT additional information to filter down and refine my original sequel query. So really quickly and let me grab that information.

Jenn Snider: I'm gonna go ahead and put it on the front. I'm telling ChatGPT, I have another table called customer address. The columns within this table are my customer ID street sitting in state. I also mentioned that my customer ID is a foreign key to the customer table. And then the previous SQL query to filter customers from Chicago. So I'm gonna go ahead and send this to chat. And of course, it's working. It's magic as usual. And so as you can see, the SQL query is pretty much the same. The only difference is that it's adding an additional join because I have an additional table. It's also using the wear clause because I asked, I wanna filter down by location. And so it's looking at, the column which has Chicago. And of course, ChatGPT is reminding me make sure to adjust the table and column names according to your actual schema if they differ. So just to pause really quickly, I've asked ChatGPT to create a sequel query for me. Okay? By being specific in giving my table names, my column names and the foreign key. And I've also asked ChatGPT to refine that query to narrow down my results by location. But let's say I have a long running query. Let's just pretend I have a long running query and I wanna make my query more efficient. ChatGPT can do that for me. So really quickly, let me grab my prompt that I have ready to go.

Jenn Snider: And I'm asking ChatGPT, I want to create one or more indexes that makes this query most efficient, please assist. And so ChatGPT is giving me some recommendations on how I can optimize and make my query more efficient by creating an index for the order table, customer table and customer address. What I love about ChatGPT is telling me at the end admin, these indexes will help the database engine efficiency, but it also says, keep in mind, while indexes can significantly enhance your query performance, they do come with some overhead in terms of storage and maintenance. So it's also telling me, hey, you can do this but also be my fault of this. So before we transition to our walk portion of today's workshop, I just wanna remind you and reiterate be mindful of ChatGPT limitations and also be mindful of the best practices that I have given. You. Again, the more precise you are in your description, the better the generated query will be. So this is just a perfect example of how ChatGPT can be your handy partner when refining your data and creating see queries. So before a transition, I'm gonna pause right there if anybody has any questions, feel free to put them in the chat, my team mate, Sheldon is moderating the chat. So any questions that you have, he'll be able to assist you?

Sheldon Spring: Yeah, really quickly. On that note, Jen, one of our attendees noted how the prompts that I was posting in the chat are not able to be copied. And pasted. Apologies for that. I just posted a link to the Google doc that houses all the prompts there. So if you would like to practice with Jen walkthrough, feel free to go to that page copy and then place them in the same order that she did.

Jenn Snider: Thank you so much Sheldon.

Jenn Snider: All right. I'm gonna go ahead and go back to my presentation. So as I said earlier, we're gonna go ahead and cover the walk portion of today's workshop. So many of you have already nailed the step of creating and refining your SQL for working with documents that are bursting with text tables and I catching charts. But the challenge shifts to efficiently analyze and drawing insights from data points. And this is where presentation enhancer will help you achieve this. Presentation enhancer is a free Google slide add on that utilizes AI to automate executive summaries or an entire slide deck or top tracks for individual slide. Now, for this next activity, we'll be generating both options for a presentation of your selection. And this is where one of the prerecord of bringing your own presentation comes into play. If you don't have your own presentation deck that's completely fine. We do have a sample one. I believe the link to that presentation deck is listed within the health documentation. So feel free to reference that and review that presentation. But before you do, there are some things that I wanna cover if you have your own presentation or utilizing the demo presentation we're providing to you. If you're using the demo presentation, download it, but download it as a copy. So that other users within this workshop is not updating the same presentation that you're utilizing, also update your upload your presentation into your Google drive as Google slide. Okay? If you have a powerpoint presentation, go ahead and upload it into your Google drive, the caveat with presentation answer, it works best with Google fly. So you would have to save your powerpoint presentation as a Google slide in order to use the presentation enhancer feature. So with that being said, I'll go ahead and switch and demo presentation enhancer for you. As I said, earlier, it is a Google add on. There is a link to presentation enhancer within the Google doc that is listed in the chat. Feel free to add this add on onto your browser. Go ahead and go through, yes, I agree with all the things so that you can utilize it today. And as I mentioned earlier, this is the demo presentation that we're utilizing I'm gonna go ahead and utilize this. I'm gonna go ahead and make a copy. I'll make a copy of the entire presentation.

Jenn Snider: I'll go ahead and rename it and make the copy.

Jenn Snider: Close that out, go ahead and close this out. All right. So here's my presentation. It's the standard quarterly business review. I mean you have, your agenda, the team executive summary, some data points, so on and so forth. So pretty general of what you want to present to your customer. But I'm having difficulty creating an executive summary or a talk track. So first, I'd like to back up a little bit and say, you know, if you're a CSM today or you're new to the organization, and you're trying to find out how data is being collected, what is valuable to the customer? And you're not quite sure how to convey that in your own voice, presentation enhancer will help you with that. So in order to do that, I'll go ahead and select extensions on the top menu, I will select presentation enhancer and select enhanced presentation. Now you'll notice a menu on the right hand side will populate, you have the option to generate an executive summary or get a talk track for the curtain current slide. So I'm gonna go ahead and select generate executive summary… and you'll notice I'll have some options here. I can select the style. I can select the persona and the format. So this is really cool because it could be whatever you pick and choose for style. I'll go ahead and select formal persona. I'll leave it at executive. But you do have some other options here. And for format, I'm gonna go ahead and select a paragraph. So once I have all of that, I'll go ahead and select generate executive summary.

Jenn Snider: And while this is populating, feel free to put any questions within the chat if you have any.

Jenn Snider: All right. So now I have my executive summary. So what presentation enhancer has done has reviewed my entire deck tool, the most valuable information and place it in a paragraph as I requested it. So as you can see, just high level simple ticket has demonstrated significant success this quarter, resolving a total of 6,600 tickets with an oppressive average resolution time of 16 hours. Pretty valuable. And as I scroll down, let's see if there is any other helpful information that I would like to highlight return on investment. I think that that's really important. This has led a substantial savings of approximately 728,000 per year equating to the full equating to the work of 19 full time employees, pretty impactful. So things to be mindful of if you feel like that this is not in your voice or you need to make some modifications that's okay. The biggest takeaway here is that AI is scanning your presentation, pulling out the most valuable data for you to deliver to your customer. And also it's gonna save you time previously. And today I create presentations. And before magic, it would take me hours to generate a good talk to. So you have the ability to copy and you can place it anywhere within your presentation. So here, I do have a slide for executive summary and I can place it under key takeaways here. On the right hand side, you can also place this information on a one pager. Let's say you wanna send out a high level overview after your T. BR. You can definitely put this information in your one pager. Now, let's go ahead and talk about a talk track… presentation. Enhancer can definitely do that for you. So I'm gonna go ahead and select the back button right here. And I'll go ahead and select the top tracked. And as you can see same options as before I'm, gonna go ahead and stick with the style of confident. I'll go ahead and select the persona of leader. And I want a bulleted list.

Jenn Snider: All right. So here's my bulleted list in the team has effectively resolved 6,600 out of 7,000 tickets received. The average resolution time stands at 16 hours per ticket. A total of 130 end users have been assigned with an estimated 91 percent of them utilizing the platform for request handling. So again, here's another example of AI looking at the talk to, the current slide and creating a talk track for you. Another thing I wanted to point out before we go to the next portion of today's workshop. You do have the ability to make your talk track or your executive summary either longer or shorter. So you do have that option.

Jenn Snider: All right. And that is presentation enhancer. It's a pretty awesome tool. I think it would give, your team or yourself confidence in presenting data to your customers and telling a Rich story.

Jenn Snider: All right. So now, we are going to talk about the run portion of today's workshop. So really quickly the two previous exercises represent ways that you can leverage AI independently of our platform. As we transition to the run portion of the workshop, we will walkthrough the dynamic content creation process to showcase how you can generate insights as part of your presentation for on step.

Jenn Snider: So let's go ahead and jump in Matik. So one second while I switch gears here, let me.

Jenn Snider: My apologies. I have a password vault.

Jenn Snider: All right. So before you log into your self serve account, you were supposed to create one prior to the workshop. If you haven't created one, that is fine, you can create a self serve account. Right now. There should be a link in the chat that will give you instructions on how to do so, utilize your Google or Gmail address. I wanna put a caveat out there. If you are a Matik customer today, you're probably already users. So you have to add the plus sign and Ai customer success workshop before the at sign in your e-mail address, in order to create a self serve account.

Jenn Snider: All right. I'm gonna pause for a couple of minutes just to make sure anyone that is following along that are able to log in.

Sheldon Spring: One thing I'll quickly note as well, everyone while Jen is pausing is when you go through the workflow to set up a self serve account, you should then receive an e-mail sent to the e-mail address that you use, whether it be your personal e-mail address or a work related e-mail address. But with the plus and the AI Customer Success workshop addendum added to it, either way, you should receive an e-mail from our CEO Nick with the title line, log in to Matik. So you're gonna locate that. And then it should allow you to access this page. And if you have any questions, please feel free to put them in the chat and monitoring that. We also have another one of our team members on standby to help with questions as well.

Jenn Snider: Thank you so much. Even so what I mentioned earlier regarding the sample data report, if you don't have an excel or Google sheet today that's fine. We do have a sample one. It should be listed within the help documentation. Once you click on the link, feel free to save a copy of the sample data. And this is what, the Google sheet looks like just for your reference. So, before we talk about AI insight, I want to connect the dots for you a little bit. So we're gonna have to go through a process in order to utilize the AI insight. And the process is we're gonna create a data source first, which is the data report that we're gonna utilize. We're then gonna create a piece of dynamic content and then we're gonna create another piece of dynamic content, which is the AI insight. And then we're gonna test it so you can see how all that flows together. So with that being said, let's go ahead and get started really quickly. I'm gonna save a copy of this.

Jenn Snider: Copy of sample data. I will go ahead and close this out. And now, I'm gonna connect my data source or create a data source. So in order to do that, if you are already logged in and if you're not that's fine. We will, this session is recorded. So you have something to reference to at your own time, you'll go ahead and select data sources and we're gonna add a new data source. Now, you have two ways of doing this. I can either select a Google sheet here on my dashboard or I can select add new. I'll then select the data source type. This case, it is a Google sheet. So I'll go ahead and scroll and find my Google sheet. And then we wanna give this data source the name. So I'll go ahead and put a… or so… and then we're gonna go ahead and select the Google sheet. So one second, just let me refresh.

Jenn Snider: All right. There we go. So here is the copy of my sample data for the workshop. We'll go ahead and select this and we'll go ahead and add the data source. Now, I have a data source. The next step is to create a piece of dynamic content. So we will select dynamic content located on the top navigation menu… and we're gonna go ahead and just select add new. So the type of dynamic content that we want to create will be data driven. This information will pull from your data source, which is the one that we just created. We're gonna go ahead and give this dynamic content a name. I'll go ahead and call it resolution time. So basically I'm pulling the info from this data source and it is tickets by day. I have two columns of channel and resolution time. So the channels are phone, social, web and e-mail so for my description, I'm just gonna type in quickest… resolution in.

Jenn Snider: All right. Now, the next thing I need to select is my data source. So it's the data source that we just connected, which is the Google sheet. I'll go ahead and select the data source and you'll notice my screen has changed. I have the option of creating a piece of dynamic content in the form of text chart table or image. I'm gonna go ahead and select chart. And then it's asking me for my sheet input and my sheet output for today's. Example. I only need an output. I wanted to output data. But I also wanted to mention that I do have the ability to select an input and what that means is if I have any calculations within my Google sheet, Matik, this platform can go ahead and pick up that calculation and output it into my dynamic content and as a result be placed onto my presentation. So it's not just a feature of taking data and placing it. It's also working with complex data to output that information as well. But since I only wanna see output, I'll go ahead and enter my column names, which is channel and resolution time.

Jenn Snider: And then I'm going to select the data I wanted to pool from my data source or my Google sheet. So I'll go ahead and highlight these two columns. And then next, before I add this piece of dynamic content, I wanna make sure it works. So I'm gonna go ahead and test this really quickly… and it works just fine. Here is my tickets by day, channel and resolution time and the data that I want to populate. So I'll go ahead and add this piece of dynamic content. And now I'm ready to utilize the AI insight. So in order to do that, we're gonna go ahead and add another piece of dynamic content. So go ahead and select add new. And this time you're gonna select the insights file. So go ahead and select create and we're gonna give, the dynamic content a name. I'll call it time… insights. So I want to give, the description as well. So I'll put the thinking quickest.

Sheldon Spring: Or…

Jenn Snider: All right. So this is where the magic happens here. Within my insight screen, I have the ability to describe if I click this drop blocks, I have more options. So if I wanted to describe, the AI insight feature will provide a description of your dynamic content output. If I were to like analyze, it, will identify trends when your dynamic content returns, complex data summarize, provides a high level summary of the data return by the dynamic content. And of course, last compare looks at two if different types of dynamic content. So you have the ability to select and choose what type of data you want to show to your customer. Again. I'll go ahead and select describe, and then you have to select the piece of dynamic content you want the AI insight to review and describe for you. So since we just created resolution, fine, go ahead and select that. It's not populating here. Let me refresh my screen really quick. I don't know if it's saved.

Jenn Snider: Fine.

Jenn Snider: Want to describe, there we go. Resolution sign. I'm gonna go ahead and select a paragraph. All right. Again, you have the option to have one sentence, a bulleted list or a number list, or a paragraph. I'm gonna select paragraph and let's go ahead and test this.

Jenn Snider: All right. So the AI in site took a look at my piece of dynamic content. It looked at the insights items that I wanted to complete and it outputted a paragraph for me. So let's go ahead and take a look at this. The provided data illustrates the resolution time across four different communication channels. The phone channel is the most effective with the quickest resolution time of 12 units. The social channel follows with the resolution time of 23 units. The web channel has a slightly longer resolution time of 32 units. The e-mail channel has the longest resolution time of all. So on and so forth. This looks good. But what's the units, why are you giving the units? And so when we look at this and we think about this, it's in minutes. I didn't put that clearly in my description in my dynamic content. So now I'm gonna go ahead and refine that. So really quickly, let me go ahead and add this piece of AI insights because I like it and I don't want to create this again. So we'll go ahead and add it. We'll go back to our original dynamic content that we created which is called resolution time and we're gonna refine this description. So this is going back to our initial exercise of crawl, right? Best practice, be as specific as possible. So you have the best results. So I'm gonna go back to my description, quick solution time by either phone, social web or e-mail. I'm gonna put measured… in minutes.

Jenn Snider: And I'll go ahead and update that. I'll go back to my dynamic content. The Ain site, I'll go ahead and test this again. So it's best practice to test and test of often. I'm showing you the process of what I would do for my customer utilizing this a site.

Jenn Snider: All right. The data states the average resolution time across four different communication channels. The phone channel offers the quickest resolution time averaging around 12 minutes. So now it's looking at my description of the piece of dynamic content that I wanted to describe and it's usually using the time and minutes instead of us. Now, as I look over, this is great. But what would my customer wanna see? I think they would wanna see their ideal resolution time and let's say that ideal resolution time is under 39. So let me go back to my initial dynamic piece of content and let me update this description again… and add the ideal.

Jenn Snider: It is.

Jenn Snider: Let me go ahead and update those. Let me go back to my inside. Let's go ahead and test this again.

Jenn Snider: All right. Data provided, the resolution time across various channels at fastest resolution time is achieved through the phone channel, averaging in four minutes. As I look through this, if I look at the last line, it's noteworthy that the ideal resolution time is targeted at under 30 minutes, a standard that is currently only met by the phone and social channels. So now I'm not only telling my customer that what is the fastest resolution time in? But I'm also pointing out that our AI is also pointing out that it realizes the targeted resolution time is under 30 minutes. And here are the two pieces of the channel that meet that goal. So when you place this piece of dynamic content into your presentation one, it saves you time from reading the data and analyzing it yourself. But it also gives you the opportunity and leverage AI to not only make you more efficient but help you understand what would be valuable to that customer.

Jenn Snider: So that is everything within this workshop. So I'm gonna back up a little bit just as a reminder of the crawl phase. We utilize ChatGPT to create a sequel query for us to refine that query as well. Those are something that you can use today. Another item that you can use today is presentation enhancer. During our walk phase, being able to create a executive summary and a talk tax of track for specific slide. And then finally the run phase utilizing the aim sit within the Matik application to review dynamic content and analyze it for me. So with that being said, let me go ahead and switch back. Any questions… please put them in the chat or Shelton. If you have any questions, let me know.

Sheldon Spring: Yeah, thank you so much, Jen. We do have one question that's coming from your lean. And they said for the presentation enhancer, plugin, what assurance as far as data security is provided? Especially when leveraging it for customer business reviews that includes metrics. So Jen… would love to hear your thoughts. I can a few of my own, but yeah, that was one of the questions we received.

Jenn Snider: For sure, for sure. So I know we do have a help documentation regarding this particular item explaining what our presentation enhancer is doing in terms of reviewing, your deck utilizing AI and what is happening with that data. I can't tell you right now verbatim what it says, my apologies, but I know we can provide that item within the help documentation. I think it's listed on there. We have a help center documentation regarding presentation answer. And on the very bottom, it lists what our process is doing. If you have any additional questions regarding that or if that particular documentation does not answer your question, that's completely fine. Please put it in the chat. We'll go ahead and notate it. And then I know when this workshop is gonna be on demand, we will, I believe we'll also reach out to any questions that we haven't answered to everyone here. So that you have it for your reference. Hilton. Did you wanna add to that?

Sheldon Spring: Yeah. I was just gonna say to Jens point, I posted two links. I'm about to post a third one. But first link is our presentation enhancer article that was written up and isn't available on our online help center. At the bottom, there's a section that talks about security and sort of privacy. I also listed just Matik, company wide privacy policy. Although that being said, we don't store any content of any, anyone that leverages presentation enhancer. We're just making it available as a means of helping customers basically expedite the process of having to create talk tracks and executive summaries. And then the last thing I'm gonna post is the open a ChatGPT for like research overview. I think our keynote speaker this morning, Lisa, she noted that ChatGPT in terms of the models that they use, I think there's a lot of security concerns around, you know, if I use ChatGPT or some other system, or platform powered by open AI, is my information? Is my customers information? They're going to be utilized to train these models? And she very promptly dispelled, that thinking, and noted that they have models that they're working off of that, do not leverage whatsoever any of the information that's being touched sort of extraneously by their tools. So I completely hear you, Arlene, it's always a question of security but hopefully that helps answer the question. And then we have.

Jenn Snider: Okay.

Sheldon Spring: Of course, of course, we have another question Jen from Winnie, who asks, say I have 50 companies I want to reach through presentations? Can it take a template and auto make those 50 presentations with dynamic insights? Do I still need to go to each presentation to make those insights? So once again, Jen would love to hear your thoughts on this one.

Jenn Snider: I got good news for you, Winnie. No. So if you are a Matik customer and you're talking about Ain sites, once you have your templates, there's a process called tagging. Once you create your dynamic contact, you would tag your templates, okay? And the input or the customer that you would select to run that presentation, Matik would know it's for X customer there for this dynamic content or AI insights is specific for that one customer. So you would have the ability to bulk generate a presentation for your 50 customers. And Matik would know based on the customers that you selected, the AI insights would be specific to those 50 customers. The only thing you would have to do is spot check the presentation to make sure that the data populated as expected. But, yeah.

Sheldon Spring: Yeah. And wine had a follow up where they asked, how do I prepare the data? Do I make 50 data sets? I think Jen touched upon this in her response, but a lot of our customers, they're working off, you know, very tables within data sources that they connect to. But normally, what we see is, and this is another facet of our platform Matik. But normally, what we see is we work with admins to create what are known as inputs and inputs can be thought of as filters essentially on a dashboard or in this case, a data source. And that tells Matik, okay, I am pulling from X tables from this database and I wanna pull information for these 50 customers that I have that would be an input in Matik. And that would be essentially how Matik is able to pull information that is specific to the accounts that you're interested in. So a little bit of nitty gritty background, into Matik as a tool. But in short, Winnie would not have to essentially make 50 data sets. You would not have to populate the data yourself that all of that process would be done in the back end by Matik. Great question. Though… any other questions from attendees, either about some of the activities that we went through or I should say Jen went through when it came to a ChatGPT, presentation enhancer, automated insights within Matik, any and all questions are welcome.

Sheldon Spring: I question, I'm sorry… quick question, Jen, how have you seen automated insights be utilized by maybe some of the accounts that you work with that Matik, you know, once Matik generates this wonderful, you know, summary or analysis or even comparison perhaps between two distinct data points or pieces of dynamic content. What, what have you seen admins then do with the insight, dynamic content? How do they incorporate it?

Jenn Snider: That's a really good question. So, for instance, I have one customer who is in the transportation space and utilized AI insights for their quarterly deck review. And basically AI insights were able to look at their data that they have stored of when their peak rides are, and when is their downtime this utilized? Or this enable the admins to look at? Okay. Well, maybe we need to run a special or a discount in the afternoon so we can get more rights or transportation requests within the afternoon. And that's not something that they had eyes into previously. So that was able, to give them insight on what their heavy traffic is, when during the month or quarter, is there a heavy traffic time of day? It just opens so much data for them to utilize. And also gave their CSM the opportunity to not only tell a story but tell a story with impactful data for their customers.

Sheldon Spring: That's amazing. And all of that is sorta automated as part of, the generation process.

Jenn Snider: All of it all, yeah, automated part of the generation process. I would have to say, the admins or, the insights team who does maintain the data. They're the ones that we work with closely to set up everything. But as an end user, your CSM team, they would just select the accounts, run the presentation and, it would be generated within minutes.

Sheldon Spring: That's magical. Amazing. I'll stop there. But thank you, Jen, for answering that question.

Jenn Snider: You're welcome. Awesome. I don't think we have any other questions so we can wrap up. Thank you so much for walking us through all of this. Everyone you'll have access to the recording. So don't worry. And yeah, I'll pass it off to get up. Awesome. Thank you so much. But the one thing that I wanted to mention before we wrap everything up first, thank you so much for your time and your interest and your curiosity on how to utilize AI to better your data and tell a better story for your customers. Also wanted to mention that the Matik presentation, an answer for Google slides is a free tool. So you can use that right now after the workshop, if you like. So please utilize that as a tool in your toolbox. Another thing I wanted to mention is your self service account as a free trial, the Matik, which is good for 30 days. So if you create dynamic content, putting your presentation, please be mindful to download your presentation before the 30 days expires. And when you log into your self service account at the top, it'll tell you trial 29 days left. But if you love Matik so much and you want to continue using Matik, feel free to reach out to our sales team for a demo, they'll be more than happy to show you how Matik can make your life easier. That's all I got.

Sheldon Spring: And final thing on my side is I'm going to post a link to this presentation for people to refer back to. If as Jen was saying, you want to reference the recording of the session, maybe follow along with the slides on your own device. So, I'm, gonna post that right now, but yeah, I just want to reiterate. Thank you everybody for your interest in turning data into insights, leveraging AI and hearing us talk about how to do so with tools that are available today, but then also some things that are specific to Matik. So thank you.

Jenn Snider: Thank you. Take care, everyone.

 

---

Click here to see all sessions & speakers, and navigate to recordings.

Join Our Newsletter

Don’t miss our next blog post & other great content!

Workshop: How to Use AI to Turn Data into Insights

Does your CS team struggle with translating data into insights? Are you trying to figure out how to get AI to do it for them (maybe even yourself)? Then you’ll want to attend this workshop, where we will practice using AI to create high-level takeaways and insights from data and its visualizations.

You’ll come away from this hands-on session with new data insights that you & your team can immediately use.

The below are not required to attend the workshop, but in order to get the most out of it:

- Bring a SQL query you’d like to optimize and work on

- Have a Google Slides or PowerPoint Presentation ready that includes data that you’d like to get insights on

- Bring a Google Sheet or Excel report of non-formatted, raw data you would want to leverage in a presentation

Speaker

Jenn Snider - Sr. Technical Account Manager at Matik

LinkedIn

Jenn Snider, a seasoned Sr. Technical Account Manager with an impressive 8-year track record. Having excelled at Yardi Systems and Atlassian in the past, she brings a wealth of expertise to her current role at Matik. With a passion for delivering top-notch customer experiences, Jenn continues to thrive in the ever-evolving tech landscape.

Watch Workshop: How to Use AI to Turn Data into Insights

‍Transcript

This transcript was created by AI. If you see any mistakes, please let us know at marketing@matik.io.

Matik MC: Awesome. Okay. Hi, everyone. Welcome to Matik, AI and customer success summit is the last session of what's been a really educational day. You are currently watching the workshop of how to use AI to turn data into insights. I'm very excited to introduce our speaker. And as I introduce her, I'm gonna have her answer our icebreaker question, which is what would you use AI to automate in your life outside of work? So we are lucky today to have Jenn Snider, senior technical account manager at Matik presenting.

Jenn Snider: Hello, everyone. Thank you so much for joining the session. I'm excited to share what we have with you today. And if I can use AI to automate a portion or something in my life that's a really good question box. I mean we're not there yet. On a funny note, blow dry, my hair would be amazing. But aside from that, I think I would have AI tell me… what grocery stores would carry sugar free bacon to sugar free and you can only find it in certain places, one whole foods, but the grocers near me, don't really have it. So if I could tell me the groceries within like a four mile radius, that would make my life easier.

Matik MC: I love that answer. But maybe instead of sugar free bacon, I might be looking for something else for me to, before I hand it off to Jen. Just a reminder for everyone watching. If you have any questions at any point in the session, go ahead and go to the Zoom control bar and use the Q and a feature. We'll be answering questions throughout the session as well as at the end. Okay. Jen, I'll hand that to you. Awesome. I will go ahead and share my screen.

Jenn Snider: All right. As I said earlier, thank you everyone for joining the session. We're excited for you to see what we have in store for you. List in front of you are the agenda items that we're going to be covering today, in addition to some time estimates of each exercise, so that you're aware of what's gonna come next and how long it's gonna take. Just as a quick reminder, make sure you are set up with the following that's an e-mail link to your Google account, a SQL query that you would like to refine if you don't have one that's okay, don't worry about that, a presentation you would like to gain insights from, and a data report either in a Google sheet or excel format. If you don't have any of these items, don't worry that's okay. We do have a sample presentation. I ready to go for you to utilize for today's exercise in addition to a sample data sheet report as well. You'll find a help documentation within the chat that has linked to everything that we're gonna utilize today, and also some reference as we walk through these exercises. Once we're done with everything, we'll go ahead and wrap up with a Q and a at the last 10 minutes of the workshop.

Jenn Snider: So at a high level, we'll be exploring some content today using the crawl walk run approach. So think of it as like taking small steps to harness the power of AI, we'll go ahead and start with the crawl phase which is something that you can get started with right now. And then we'll go ahead and transition into walk which is a bit more advanced but still very doable today. And then finally we'll reach the run stage where you'll see the full potential of AI in action. So let's go ahead and kick off with the crawl phase. And in order to complete this, we'll be utilizing ChatGPT to create and refine a query. So ChatGPT is an AI powered chat box that has gained significant attention for the sheer variety of comments and queries that it can respond to and provide inputs for. So, before I go ahead and switch gears and log into ChatGPT to start the first exercise, go ahead and log into your ChatGPT account. Now, if you don't have one today, you can create one, it is free. But once you log in, we're gonna go ahead and start with either option one or option two. So if you don't have a SQL query, we can, we're gonna leverage ChatGPT to create one for us. If you do have a SQL query and want to refine it and see what I do to refine a SQL query, you can do that as well. But for today's, workshop, I'm gonna go ahead and do both. So let me go ahead and switch my window. I'm gonna wait. I'll say like a minute just to make sure that everybody can log in to ChatGPT, so you're able to follow log. And while we're waiting for this one minute, if you have any questions or excitement that you want to share, go ahead and put it in the chat.

Jenn Snider: All right. So before I get started with leveraging ChatGPT to create a sequel query for myself, there's some things that I want to bring to your attention. And that's actually the limitations of ChatGPT. A couple of the limitations that I really want to highlight is the inability to ask clarifying questions. So if your input is ambiguous ChatGPT, my guess your intention rather than asking clarifying questions for a better understanding. So please be mica of that. Another thing I want to bring to your attention is the limited context window that CPT. The model itself has a fixed context window which means they may not remember information mentioned earlier, if you have a longer conversation which can lead to inconsistent responses. So please keep that in mind. With those two points brought up, I would like to highlight some best practices that will help you in leveraging ChatGPT. First be very specific, especially when you want a ChatGPT to create a query for you. And by specific, I mean list the table and call in names to avoid ambiguity, also clearly state any conditions, filters or sorting requirements. And lastly, the more precise you are in your description, the better the generated query will be okay. So with all those points in mind, let's go ahead and dig in and get started. Now, I have some prompts here that are already available just to speed things up because I am a slow typer, nobody wants to watch me type. So I'm gonna go ahead and paste in my first prompt and based call. I'm using a very easy example of customer orders. So what I mean by customer orders is I have an order table and a customer table within my order table, my column ID'S are my product and my price. And with my customer table, my column ID names are name and e-mail. I also let ChatGPT know that my customer ID and my order tables is my foreign key referencing the ID column and my customer table. So after I laid out all the tables and columns that I have, I'm asking ChatGPT to write a sequel query that's gonna find the total revenue of each customer by calculating the sum of their order prices. So now that I have that in the chat, I'll go ahead and send it.

Jenn Snider: And ChatGPT just created a simple query for me. So as we take a look at this, ChatGPT is utilizing the join and group by clause to make this happen. As we scroll down a little bit and take a look at the query and see what ChatGPT is doing. They're selecting the ID name and e-mail columns from my customer table, and also calculating the price utilizing the some function. It is also using a join to join the two tables because I referenced my foreign key and both my orders and customer table. And it's grouping by the results, the name and e-mail as you can see, it's saying here, just replace customer ID name and e-mail with the actual column names in your database. If they are named differently, you can then copy this code, put it in your database for or whatever tool that you're utilizing to search for those results, which is really cool. But let's say I wanna take this a step further. I have the information that I want but I wanna filter it down by customer location. The one caveat to that is I've already told ChatGPT my order and customer table which doesn't include location. So I'm gonna have to tell a CPT additional information to filter down and refine my original sequel query. So really quickly and let me grab that information.

Jenn Snider: I'm gonna go ahead and put it on the front. I'm telling ChatGPT, I have another table called customer address. The columns within this table are my customer ID street sitting in state. I also mentioned that my customer ID is a foreign key to the customer table. And then the previous SQL query to filter customers from Chicago. So I'm gonna go ahead and send this to chat. And of course, it's working. It's magic as usual. And so as you can see, the SQL query is pretty much the same. The only difference is that it's adding an additional join because I have an additional table. It's also using the wear clause because I asked, I wanna filter down by location. And so it's looking at, the column which has Chicago. And of course, ChatGPT is reminding me make sure to adjust the table and column names according to your actual schema if they differ. So just to pause really quickly, I've asked ChatGPT to create a sequel query for me. Okay? By being specific in giving my table names, my column names and the foreign key. And I've also asked ChatGPT to refine that query to narrow down my results by location. But let's say I have a long running query. Let's just pretend I have a long running query and I wanna make my query more efficient. ChatGPT can do that for me. So really quickly, let me grab my prompt that I have ready to go.

Jenn Snider: And I'm asking ChatGPT, I want to create one or more indexes that makes this query most efficient, please assist. And so ChatGPT is giving me some recommendations on how I can optimize and make my query more efficient by creating an index for the order table, customer table and customer address. What I love about ChatGPT is telling me at the end admin, these indexes will help the database engine efficiency, but it also says, keep in mind, while indexes can significantly enhance your query performance, they do come with some overhead in terms of storage and maintenance. So it's also telling me, hey, you can do this but also be my fault of this. So before we transition to our walk portion of today's workshop, I just wanna remind you and reiterate be mindful of ChatGPT limitations and also be mindful of the best practices that I have given. You. Again, the more precise you are in your description, the better the generated query will be. So this is just a perfect example of how ChatGPT can be your handy partner when refining your data and creating see queries. So before a transition, I'm gonna pause right there if anybody has any questions, feel free to put them in the chat, my team mate, Sheldon is moderating the chat. So any questions that you have, he'll be able to assist you?

Sheldon Spring: Yeah, really quickly. On that note, Jen, one of our attendees noted how the prompts that I was posting in the chat are not able to be copied. And pasted. Apologies for that. I just posted a link to the Google doc that houses all the prompts there. So if you would like to practice with Jen walkthrough, feel free to go to that page copy and then place them in the same order that she did.

Jenn Snider: Thank you so much Sheldon.

Jenn Snider: All right. I'm gonna go ahead and go back to my presentation. So as I said earlier, we're gonna go ahead and cover the walk portion of today's workshop. So many of you have already nailed the step of creating and refining your SQL for working with documents that are bursting with text tables and I catching charts. But the challenge shifts to efficiently analyze and drawing insights from data points. And this is where presentation enhancer will help you achieve this. Presentation enhancer is a free Google slide add on that utilizes AI to automate executive summaries or an entire slide deck or top tracks for individual slide. Now, for this next activity, we'll be generating both options for a presentation of your selection. And this is where one of the prerecord of bringing your own presentation comes into play. If you don't have your own presentation deck that's completely fine. We do have a sample one. I believe the link to that presentation deck is listed within the health documentation. So feel free to reference that and review that presentation. But before you do, there are some things that I wanna cover if you have your own presentation or utilizing the demo presentation we're providing to you. If you're using the demo presentation, download it, but download it as a copy. So that other users within this workshop is not updating the same presentation that you're utilizing, also update your upload your presentation into your Google drive as Google slide. Okay? If you have a powerpoint presentation, go ahead and upload it into your Google drive, the caveat with presentation answer, it works best with Google fly. So you would have to save your powerpoint presentation as a Google slide in order to use the presentation enhancer feature. So with that being said, I'll go ahead and switch and demo presentation enhancer for you. As I said, earlier, it is a Google add on. There is a link to presentation enhancer within the Google doc that is listed in the chat. Feel free to add this add on onto your browser. Go ahead and go through, yes, I agree with all the things so that you can utilize it today. And as I mentioned earlier, this is the demo presentation that we're utilizing I'm gonna go ahead and utilize this. I'm gonna go ahead and make a copy. I'll make a copy of the entire presentation.

Jenn Snider: I'll go ahead and rename it and make the copy.

Jenn Snider: Close that out, go ahead and close this out. All right. So here's my presentation. It's the standard quarterly business review. I mean you have, your agenda, the team executive summary, some data points, so on and so forth. So pretty general of what you want to present to your customer. But I'm having difficulty creating an executive summary or a talk track. So first, I'd like to back up a little bit and say, you know, if you're a CSM today or you're new to the organization, and you're trying to find out how data is being collected, what is valuable to the customer? And you're not quite sure how to convey that in your own voice, presentation enhancer will help you with that. So in order to do that, I'll go ahead and select extensions on the top menu, I will select presentation enhancer and select enhanced presentation. Now you'll notice a menu on the right hand side will populate, you have the option to generate an executive summary or get a talk track for the curtain current slide. So I'm gonna go ahead and select generate executive summary… and you'll notice I'll have some options here. I can select the style. I can select the persona and the format. So this is really cool because it could be whatever you pick and choose for style. I'll go ahead and select formal persona. I'll leave it at executive. But you do have some other options here. And for format, I'm gonna go ahead and select a paragraph. So once I have all of that, I'll go ahead and select generate executive summary.

Jenn Snider: And while this is populating, feel free to put any questions within the chat if you have any.

Jenn Snider: All right. So now I have my executive summary. So what presentation enhancer has done has reviewed my entire deck tool, the most valuable information and place it in a paragraph as I requested it. So as you can see, just high level simple ticket has demonstrated significant success this quarter, resolving a total of 6,600 tickets with an oppressive average resolution time of 16 hours. Pretty valuable. And as I scroll down, let's see if there is any other helpful information that I would like to highlight return on investment. I think that that's really important. This has led a substantial savings of approximately 728,000 per year equating to the full equating to the work of 19 full time employees, pretty impactful. So things to be mindful of if you feel like that this is not in your voice or you need to make some modifications that's okay. The biggest takeaway here is that AI is scanning your presentation, pulling out the most valuable data for you to deliver to your customer. And also it's gonna save you time previously. And today I create presentations. And before magic, it would take me hours to generate a good talk to. So you have the ability to copy and you can place it anywhere within your presentation. So here, I do have a slide for executive summary and I can place it under key takeaways here. On the right hand side, you can also place this information on a one pager. Let's say you wanna send out a high level overview after your T. BR. You can definitely put this information in your one pager. Now, let's go ahead and talk about a talk track… presentation. Enhancer can definitely do that for you. So I'm gonna go ahead and select the back button right here. And I'll go ahead and select the top tracked. And as you can see same options as before I'm, gonna go ahead and stick with the style of confident. I'll go ahead and select the persona of leader. And I want a bulleted list.

Jenn Snider: All right. So here's my bulleted list in the team has effectively resolved 6,600 out of 7,000 tickets received. The average resolution time stands at 16 hours per ticket. A total of 130 end users have been assigned with an estimated 91 percent of them utilizing the platform for request handling. So again, here's another example of AI looking at the talk to, the current slide and creating a talk track for you. Another thing I wanted to point out before we go to the next portion of today's workshop. You do have the ability to make your talk track or your executive summary either longer or shorter. So you do have that option.

Jenn Snider: All right. And that is presentation enhancer. It's a pretty awesome tool. I think it would give, your team or yourself confidence in presenting data to your customers and telling a Rich story.

Jenn Snider: All right. So now, we are going to talk about the run portion of today's workshop. So really quickly the two previous exercises represent ways that you can leverage AI independently of our platform. As we transition to the run portion of the workshop, we will walkthrough the dynamic content creation process to showcase how you can generate insights as part of your presentation for on step.

Jenn Snider: So let's go ahead and jump in Matik. So one second while I switch gears here, let me.

Jenn Snider: My apologies. I have a password vault.

Jenn Snider: All right. So before you log into your self serve account, you were supposed to create one prior to the workshop. If you haven't created one, that is fine, you can create a self serve account. Right now. There should be a link in the chat that will give you instructions on how to do so, utilize your Google or Gmail address. I wanna put a caveat out there. If you are a Matik customer today, you're probably already users. So you have to add the plus sign and Ai customer success workshop before the at sign in your e-mail address, in order to create a self serve account.

Jenn Snider: All right. I'm gonna pause for a couple of minutes just to make sure anyone that is following along that are able to log in.

Sheldon Spring: One thing I'll quickly note as well, everyone while Jen is pausing is when you go through the workflow to set up a self serve account, you should then receive an e-mail sent to the e-mail address that you use, whether it be your personal e-mail address or a work related e-mail address. But with the plus and the AI Customer Success workshop addendum added to it, either way, you should receive an e-mail from our CEO Nick with the title line, log in to Matik. So you're gonna locate that. And then it should allow you to access this page. And if you have any questions, please feel free to put them in the chat and monitoring that. We also have another one of our team members on standby to help with questions as well.

Jenn Snider: Thank you so much. Even so what I mentioned earlier regarding the sample data report, if you don't have an excel or Google sheet today that's fine. We do have a sample one. It should be listed within the help documentation. Once you click on the link, feel free to save a copy of the sample data. And this is what, the Google sheet looks like just for your reference. So, before we talk about AI insight, I want to connect the dots for you a little bit. So we're gonna have to go through a process in order to utilize the AI insight. And the process is we're gonna create a data source first, which is the data report that we're gonna utilize. We're then gonna create a piece of dynamic content and then we're gonna create another piece of dynamic content, which is the AI insight. And then we're gonna test it so you can see how all that flows together. So with that being said, let's go ahead and get started really quickly. I'm gonna save a copy of this.

Jenn Snider: Copy of sample data. I will go ahead and close this out. And now, I'm gonna connect my data source or create a data source. So in order to do that, if you are already logged in and if you're not that's fine. We will, this session is recorded. So you have something to reference to at your own time, you'll go ahead and select data sources and we're gonna add a new data source. Now, you have two ways of doing this. I can either select a Google sheet here on my dashboard or I can select add new. I'll then select the data source type. This case, it is a Google sheet. So I'll go ahead and scroll and find my Google sheet. And then we wanna give this data source the name. So I'll go ahead and put a… or so… and then we're gonna go ahead and select the Google sheet. So one second, just let me refresh.

Jenn Snider: All right. There we go. So here is the copy of my sample data for the workshop. We'll go ahead and select this and we'll go ahead and add the data source. Now, I have a data source. The next step is to create a piece of dynamic content. So we will select dynamic content located on the top navigation menu… and we're gonna go ahead and just select add new. So the type of dynamic content that we want to create will be data driven. This information will pull from your data source, which is the one that we just created. We're gonna go ahead and give this dynamic content a name. I'll go ahead and call it resolution time. So basically I'm pulling the info from this data source and it is tickets by day. I have two columns of channel and resolution time. So the channels are phone, social, web and e-mail so for my description, I'm just gonna type in quickest… resolution in.

Jenn Snider: All right. Now, the next thing I need to select is my data source. So it's the data source that we just connected, which is the Google sheet. I'll go ahead and select the data source and you'll notice my screen has changed. I have the option of creating a piece of dynamic content in the form of text chart table or image. I'm gonna go ahead and select chart. And then it's asking me for my sheet input and my sheet output for today's. Example. I only need an output. I wanted to output data. But I also wanted to mention that I do have the ability to select an input and what that means is if I have any calculations within my Google sheet, Matik, this platform can go ahead and pick up that calculation and output it into my dynamic content and as a result be placed onto my presentation. So it's not just a feature of taking data and placing it. It's also working with complex data to output that information as well. But since I only wanna see output, I'll go ahead and enter my column names, which is channel and resolution time.

Jenn Snider: And then I'm going to select the data I wanted to pool from my data source or my Google sheet. So I'll go ahead and highlight these two columns. And then next, before I add this piece of dynamic content, I wanna make sure it works. So I'm gonna go ahead and test this really quickly… and it works just fine. Here is my tickets by day, channel and resolution time and the data that I want to populate. So I'll go ahead and add this piece of dynamic content. And now I'm ready to utilize the AI insight. So in order to do that, we're gonna go ahead and add another piece of dynamic content. So go ahead and select add new. And this time you're gonna select the insights file. So go ahead and select create and we're gonna give, the dynamic content a name. I'll call it time… insights. So I want to give, the description as well. So I'll put the thinking quickest.

Sheldon Spring: Or…

Jenn Snider: All right. So this is where the magic happens here. Within my insight screen, I have the ability to describe if I click this drop blocks, I have more options. So if I wanted to describe, the AI insight feature will provide a description of your dynamic content output. If I were to like analyze, it, will identify trends when your dynamic content returns, complex data summarize, provides a high level summary of the data return by the dynamic content. And of course, last compare looks at two if different types of dynamic content. So you have the ability to select and choose what type of data you want to show to your customer. Again. I'll go ahead and select describe, and then you have to select the piece of dynamic content you want the AI insight to review and describe for you. So since we just created resolution, fine, go ahead and select that. It's not populating here. Let me refresh my screen really quick. I don't know if it's saved.

Jenn Snider: Fine.

Jenn Snider: Want to describe, there we go. Resolution sign. I'm gonna go ahead and select a paragraph. All right. Again, you have the option to have one sentence, a bulleted list or a number list, or a paragraph. I'm gonna select paragraph and let's go ahead and test this.

Jenn Snider: All right. So the AI in site took a look at my piece of dynamic content. It looked at the insights items that I wanted to complete and it outputted a paragraph for me. So let's go ahead and take a look at this. The provided data illustrates the resolution time across four different communication channels. The phone channel is the most effective with the quickest resolution time of 12 units. The social channel follows with the resolution time of 23 units. The web channel has a slightly longer resolution time of 32 units. The e-mail channel has the longest resolution time of all. So on and so forth. This looks good. But what's the units, why are you giving the units? And so when we look at this and we think about this, it's in minutes. I didn't put that clearly in my description in my dynamic content. So now I'm gonna go ahead and refine that. So really quickly, let me go ahead and add this piece of AI insights because I like it and I don't want to create this again. So we'll go ahead and add it. We'll go back to our original dynamic content that we created which is called resolution time and we're gonna refine this description. So this is going back to our initial exercise of crawl, right? Best practice, be as specific as possible. So you have the best results. So I'm gonna go back to my description, quick solution time by either phone, social web or e-mail. I'm gonna put measured… in minutes.

Jenn Snider: And I'll go ahead and update that. I'll go back to my dynamic content. The Ain site, I'll go ahead and test this again. So it's best practice to test and test of often. I'm showing you the process of what I would do for my customer utilizing this a site.

Jenn Snider: All right. The data states the average resolution time across four different communication channels. The phone channel offers the quickest resolution time averaging around 12 minutes. So now it's looking at my description of the piece of dynamic content that I wanted to describe and it's usually using the time and minutes instead of us. Now, as I look over, this is great. But what would my customer wanna see? I think they would wanna see their ideal resolution time and let's say that ideal resolution time is under 39. So let me go back to my initial dynamic piece of content and let me update this description again… and add the ideal.

Jenn Snider: It is.

Jenn Snider: Let me go ahead and update those. Let me go back to my inside. Let's go ahead and test this again.

Jenn Snider: All right. Data provided, the resolution time across various channels at fastest resolution time is achieved through the phone channel, averaging in four minutes. As I look through this, if I look at the last line, it's noteworthy that the ideal resolution time is targeted at under 30 minutes, a standard that is currently only met by the phone and social channels. So now I'm not only telling my customer that what is the fastest resolution time in? But I'm also pointing out that our AI is also pointing out that it realizes the targeted resolution time is under 30 minutes. And here are the two pieces of the channel that meet that goal. So when you place this piece of dynamic content into your presentation one, it saves you time from reading the data and analyzing it yourself. But it also gives you the opportunity and leverage AI to not only make you more efficient but help you understand what would be valuable to that customer.

Jenn Snider: So that is everything within this workshop. So I'm gonna back up a little bit just as a reminder of the crawl phase. We utilize ChatGPT to create a sequel query for us to refine that query as well. Those are something that you can use today. Another item that you can use today is presentation enhancer. During our walk phase, being able to create a executive summary and a talk tax of track for specific slide. And then finally the run phase utilizing the aim sit within the Matik application to review dynamic content and analyze it for me. So with that being said, let me go ahead and switch back. Any questions… please put them in the chat or Shelton. If you have any questions, let me know.

Sheldon Spring: Yeah, thank you so much, Jen. We do have one question that's coming from your lean. And they said for the presentation enhancer, plugin, what assurance as far as data security is provided? Especially when leveraging it for customer business reviews that includes metrics. So Jen… would love to hear your thoughts. I can a few of my own, but yeah, that was one of the questions we received.

Jenn Snider: For sure, for sure. So I know we do have a help documentation regarding this particular item explaining what our presentation enhancer is doing in terms of reviewing, your deck utilizing AI and what is happening with that data. I can't tell you right now verbatim what it says, my apologies, but I know we can provide that item within the help documentation. I think it's listed on there. We have a help center documentation regarding presentation answer. And on the very bottom, it lists what our process is doing. If you have any additional questions regarding that or if that particular documentation does not answer your question, that's completely fine. Please put it in the chat. We'll go ahead and notate it. And then I know when this workshop is gonna be on demand, we will, I believe we'll also reach out to any questions that we haven't answered to everyone here. So that you have it for your reference. Hilton. Did you wanna add to that?

Sheldon Spring: Yeah. I was just gonna say to Jens point, I posted two links. I'm about to post a third one. But first link is our presentation enhancer article that was written up and isn't available on our online help center. At the bottom, there's a section that talks about security and sort of privacy. I also listed just Matik, company wide privacy policy. Although that being said, we don't store any content of any, anyone that leverages presentation enhancer. We're just making it available as a means of helping customers basically expedite the process of having to create talk tracks and executive summaries. And then the last thing I'm gonna post is the open a ChatGPT for like research overview. I think our keynote speaker this morning, Lisa, she noted that ChatGPT in terms of the models that they use, I think there's a lot of security concerns around, you know, if I use ChatGPT or some other system, or platform powered by open AI, is my information? Is my customers information? They're going to be utilized to train these models? And she very promptly dispelled, that thinking, and noted that they have models that they're working off of that, do not leverage whatsoever any of the information that's being touched sort of extraneously by their tools. So I completely hear you, Arlene, it's always a question of security but hopefully that helps answer the question. And then we have.

Jenn Snider: Okay.

Sheldon Spring: Of course, of course, we have another question Jen from Winnie, who asks, say I have 50 companies I want to reach through presentations? Can it take a template and auto make those 50 presentations with dynamic insights? Do I still need to go to each presentation to make those insights? So once again, Jen would love to hear your thoughts on this one.

Jenn Snider: I got good news for you, Winnie. No. So if you are a Matik customer and you're talking about Ain sites, once you have your templates, there's a process called tagging. Once you create your dynamic contact, you would tag your templates, okay? And the input or the customer that you would select to run that presentation, Matik would know it's for X customer there for this dynamic content or AI insights is specific for that one customer. So you would have the ability to bulk generate a presentation for your 50 customers. And Matik would know based on the customers that you selected, the AI insights would be specific to those 50 customers. The only thing you would have to do is spot check the presentation to make sure that the data populated as expected. But, yeah.

Sheldon Spring: Yeah. And wine had a follow up where they asked, how do I prepare the data? Do I make 50 data sets? I think Jen touched upon this in her response, but a lot of our customers, they're working off, you know, very tables within data sources that they connect to. But normally, what we see is, and this is another facet of our platform Matik. But normally, what we see is we work with admins to create what are known as inputs and inputs can be thought of as filters essentially on a dashboard or in this case, a data source. And that tells Matik, okay, I am pulling from X tables from this database and I wanna pull information for these 50 customers that I have that would be an input in Matik. And that would be essentially how Matik is able to pull information that is specific to the accounts that you're interested in. So a little bit of nitty gritty background, into Matik as a tool. But in short, Winnie would not have to essentially make 50 data sets. You would not have to populate the data yourself that all of that process would be done in the back end by Matik. Great question. Though… any other questions from attendees, either about some of the activities that we went through or I should say Jen went through when it came to a ChatGPT, presentation enhancer, automated insights within Matik, any and all questions are welcome.

Sheldon Spring: I question, I'm sorry… quick question, Jen, how have you seen automated insights be utilized by maybe some of the accounts that you work with that Matik, you know, once Matik generates this wonderful, you know, summary or analysis or even comparison perhaps between two distinct data points or pieces of dynamic content. What, what have you seen admins then do with the insight, dynamic content? How do they incorporate it?

Jenn Snider: That's a really good question. So, for instance, I have one customer who is in the transportation space and utilized AI insights for their quarterly deck review. And basically AI insights were able to look at their data that they have stored of when their peak rides are, and when is their downtime this utilized? Or this enable the admins to look at? Okay. Well, maybe we need to run a special or a discount in the afternoon so we can get more rights or transportation requests within the afternoon. And that's not something that they had eyes into previously. So that was able, to give them insight on what their heavy traffic is, when during the month or quarter, is there a heavy traffic time of day? It just opens so much data for them to utilize. And also gave their CSM the opportunity to not only tell a story but tell a story with impactful data for their customers.

Sheldon Spring: That's amazing. And all of that is sorta automated as part of, the generation process.

Jenn Snider: All of it all, yeah, automated part of the generation process. I would have to say, the admins or, the insights team who does maintain the data. They're the ones that we work with closely to set up everything. But as an end user, your CSM team, they would just select the accounts, run the presentation and, it would be generated within minutes.

Sheldon Spring: That's magical. Amazing. I'll stop there. But thank you, Jen, for answering that question.

Jenn Snider: You're welcome. Awesome. I don't think we have any other questions so we can wrap up. Thank you so much for walking us through all of this. Everyone you'll have access to the recording. So don't worry. And yeah, I'll pass it off to get up. Awesome. Thank you so much. But the one thing that I wanted to mention before we wrap everything up first, thank you so much for your time and your interest and your curiosity on how to utilize AI to better your data and tell a better story for your customers. Also wanted to mention that the Matik presentation, an answer for Google slides is a free tool. So you can use that right now after the workshop, if you like. So please utilize that as a tool in your toolbox. Another thing I wanted to mention is your self service account as a free trial, the Matik, which is good for 30 days. So if you create dynamic content, putting your presentation, please be mindful to download your presentation before the 30 days expires. And when you log into your self service account at the top, it'll tell you trial 29 days left. But if you love Matik so much and you want to continue using Matik, feel free to reach out to our sales team for a demo, they'll be more than happy to show you how Matik can make your life easier. That's all I got.

Sheldon Spring: And final thing on my side is I'm going to post a link to this presentation for people to refer back to. If as Jen was saying, you want to reference the recording of the session, maybe follow along with the slides on your own device. So, I'm, gonna post that right now, but yeah, I just want to reiterate. Thank you everybody for your interest in turning data into insights, leveraging AI and hearing us talk about how to do so with tools that are available today, but then also some things that are specific to Matik. So thank you.

Jenn Snider: Thank you. Take care, everyone.

 

---

Click here to see all sessions & speakers, and navigate to recordings.

Next
Request DemoSchedule Demo

Related Blogs

AICSS
See All Sessions & Speakers
Bex Sekar
  -  
August 30, 2023
AICSS
AI-powered CS: The Ultimate Guide for CS Leaders (AI + Customer Success Summit Session)
Bex Sekar
  -  
August 14, 2023
AICSS
Workshop: How to Use AI to Create & Repurpose Content (AI + Customer Success Summit Session)
Bex Sekar
  -  
August 9, 2023
© 2023 Matik, Inc.
Product
How It WorksIntegrationsAI FeaturesSecurity
Resources
BlogTemplatesAI + CS Resource HubCase StudiesHelp Center