Name: AI that Works for You: Quickbase Intelligence Package
Uploaded: 2026-05-07T17:33:21.171Z
Duration: 57 min 16 s
Description: AI that Works for You: Quickbase Intelligence Package

Transcript for "AI that Works for You: Quickbase Intelligence Package": Alright. Hello, everybody. Quickbase customers, prospects. We are super excited to talk to you today about our new product offering that we just recently launched. We're calling it the Quickbase Intelligence Package. All things AI to help you be more proficient and productive in your day. I'm Joe. I'm in product ops. I'm joined by my esteemed colleagues, Alex Peterson, James Jawaioli. Both are in the solution consultant or the forward deployed engineer space. They you're gonna probably probably interact with them a lot of them before. They exist at Quickbase to help make our product super special for you to make sure it works for your specific use cases. So I'm excited for them to really do the meat of this, presentation today in terms of demo and show you what's possible so you can get a better sense of the opportunity that that you can that that you can have with this new offering. Alright. Today, we'll kind of quickly go through our AI journey at Quickbase, talk about this intelligence package, and then some of the ROI success stories, and also how to take out a trial in the process, and then we'll go into demos, which is where I wanna spend most of our time. Cool. Alright. First, AI journey, high level. A lot of you have been with us for some time. You'll probably remember our smart builder that released a few years ago. What was the idea there? Was it just the feature we released? No. It was honestly part of a bigger plan as we tried to infuse AI across our platform to make the build management analytics, and overall data management process easier for you and all govern across our secure platform. We started out with smart builder. Now what we have done is we have enhanced the most capabilities. We've released an AI agent, which we'll talk about in a second, that allows you to make app modifications, conduct deep analysis, and all with what? Actually, very basic text. Right? You don't have to know advanced SQL or do any advanced analysis and spreadsheets. You can just ask each app bot and get deep insights on the information within your product. And you can also build formulas and pipelines. Right? A lot of those things that have been off limits for a lot of folks unless they had deep, like, based expertise. And all of this has just been a part of just making our product easier to use for you and also giving you deeper insights into everything you have and also to deliver automations, right, for some of the tedious work that you do on a regular basis. We're gonna try to take that away from you to reduce manual, entry and also, truthfully, to get rid of a lot of errors. And again, big thing at the bottom here, the the customization that the governance. Right? We wanna stay core to that so you know everything's secure. Alright. Intelligence package. What is it? So it's essentially a collection of six features that all work together within this package. It they all work to help turn unstructured and unstructured data within Quickbase, saves you ton of time by a lot of manual time consuming efforts. It delivers you insights truthfully in seconds and it gets rid of, like, a lot of manual effort. Puts things in automation for you. So that's exciting. Couple of things to highlight. The AI control center, that is like the governance capability. Right? James is gonna go into that today. And, essentially, what it is allows you to get very specific in who can access what AI capability, when, and even do different permissions at the user and group and team level. So that would you will get some insights into that today. There's this app intelligence capability that helps you understand everything about your app from the structure tables, the, pipelines and everything that's there. So that if you inherit it, you better understand it, but also that you can better make modifications and even produce copies of it later if you wanted to make something different. The AI agent, I I truthfully like this is my one of my favorite capabilities. I think of it as just chatty bb infused in the our product itself. But it understands the context of your apps, and it sits in our secure platform and allows you to do things you just couldn't before without having deep knowledge. The actions, James and I have to talk about some of this. This is about automations. Right? Extracting unstructured data from, like, emails, PDFs, attachments via OCRs. Caches via OCR. This is exciting stuff. Getting them to our platform, and then you can automatically route it in different areas. Right? So this is this is, this is some game changing stuff. Data analyzers all about predictions. Right? Predicting outcomes in the future, and, you know, actually helping your business perform better. And then there's a knowledge layer. That piece is not out yet. We won't have a demo for you today. But think of that as better making the agent all these capabilities better tailored to your business so that you actually get better on so that so that the product better understands your business and actually can produce better outcomes for you and just kinda getting trained on on what it is that you need. So that will be coming in in the next, probably in the next few quarters. Alright. Here's a few images team wanted to share. Jason and I was gonna dive into all this, kinda what this looks like, the control center, to the agent, to the analyzer. Just couldn't help myself. Just wanted to share a few images. Alright. Problems and use cases solved. Team, I I I alluded to this just a few slides ago. Just what you're able to do from an actions perspective. Like, these are very specific things, like automated invoice processing or risk classifications. Right? Imagine, like, being able to adjust documents. It already understands kind of the risks and is able to flag different things for different users, right, in different admins across the platform. Asia can help you with app modifications or very deep operational insights. It's very key. And I mentioned the app intelligence really helping with onboarding, understanding apps, and then the predictive analytics capabilities. Like, if you can think of use cases like budget forecasting. So a lot of things that could really help all of you regardless of what you do on Quickbase because we know a lot of you do a lot here. Alright. We did run a beta process team. It was very involved. It was about two and a half or so months. And the reason we did that was to hear from all of you. We heard a lot of feedback. We've learned a little bit more about how to quickly connect with you on value and to make sure that we could set this up for your success, you know, with something James and Alex will talk to in a moment. We also learned about some product apps, and these are things that are coming out. We just released the ability to copy your chats and share them with folks. And what we're gonna get to, chat history and deletion, file attachments, traceability. This is making this product even more robust so that you can come back to actually updating analysis more regularly, more easily without having to actually completely rewrite something like you've had to do before. And also you get some things like prompt suggestions and having the AI actually generate charts. And then I mentioned below here around the the knowledge layer. Right? And that coming out and being even more helping be more robust and figure out to your needs. So these are all things that we are doing because we're directly listening to you. So continue to give us a thumbs up, thumbs down icon, or even just submit your feedback directly. It's very important to us and we're incorporating into our road map. The only other thing I wanted to mention here too is just improve workflow building experiences, making actions and interacting those pipelines even easier. Right? So you can set up automations much faster. Quickly, let's talk about ROI. Where does ROI come from? Hopefully, it kind of alluded to it earlier in the conversation, but there's a lot here. Time savings. Some of this analysis could save you thirty minutes. Imagine having to do some of this analysis multiple times a day across a team many weeks in a year. This is tons of savings over a long period of time. And so just some hypothetical ideas, and I I, like, where you can actually save money. I'm using this package. Cost avoidance, this could help you potentially avoid bidding on, like, some sort of, like, or bidding from or a project from some faulty vendor or or even revenue acceleration, being able to get an additional project in for the year because you're able to operate more efficiently. Right? So these are some big things, and we'll allude to some of this, right, in some of the demos that we have. And and even James and Alex, you all have been doing a few things. Right? Maybe you can speak to some of the stuff that you've heard. Yeah. I mean, that first quote right there, it really came down to a customer was, we were working on on building a formula within their application, and it was a really complex rich text formula. You'll see some examples later, but, basically, developed a really awesome formula within maybe thirty seconds or so, and maybe not even that long. And he was just amazed that it like, for him to do that, it would have taken, I don't know, maybe an hour or so, going line by line, making sure all the syntax is absolutely perfect. Now you can just do use some natural language and boom, you have an awesome formula right there for you. Yeah. I I have something very similar, James. My my kind of comment here is, you know, this is gonna save my team ten to fifteen hours each week. I think we did actually later on figure out it was it was gonna be even larger savings. But where that kind of moment came for this this customer was when they realized that using our AI actions and actually parsing through documents and getting that information within those documents filled into a quick based record is really, a massive time saving for for them and their teams that they work with. Yep. There's a lot more out there. We just don't we could be we don't have time for it for it all. But those are some ones that just jump out to us. There's many more to come. We want you all to be a part of that story, if you know you will. Alright. If you're interested in this, just wanna hit upfront. You can take a trial self serve. Huge thanks to our billing team who put this at launch that we did. We launched this product in in the April. You're at if you're in realm admin, you can take out a quick base trial, and then you'll also be able to see in your console, like, how many days are left in the process. If you're not a realm admin, just contact, your realm admin to to to be able to get access to this. And then, obviously, with the governance capability, you have to control who gets access to what and when so it doesn't overwhelm users. But we understand to see value, you really gotta kinda play around with it. And that's what we're trying to do here with you can do that with your rep, yourself, and then even with with some of our esteemed colleagues here on the call. Alright. With that, why don't we just kick off with demos? James, Alex, I'll I'll pass it over to you. Stop sharing. So I think I'll I'll kick it off here. So, actually, I think that's a good segue. I'm gonna start over at the control center here. So this is within the admin console. And what you're looking at right here, this is the platform level AI controls. So you can really control every different type of function, that is available within our AI capabilities here. So different ways that you can build applications through AI, different ways you can interact with the AI agent. So, you know, you can create and update records, you can build with it, you can obviously use it to, query data, summarize data. So depending on if you want anyone within your platform to be able to do that, you can set that right here. Additionally, there's there's other, you know, analyzation, type features within here, that you can turn on and off as well. And you can see, what's directly within the intelligence package or what might be in beta as well. So that's all available for you to govern. Again, nice little note up here at the top. Quickbase never trains the AI models on your data and your data is not retained by any Quickbase AI technology providers. So I just wanna call that out here as well. It's pretty important for for a lot of you, I'm sure. But once you determine who, like, what capabilities you just want anyone on on the realm to be able to do, You can take that a step further so if you go to your permissions page. It's going to be governed very much like you can with other pipelines channels if you're a raw man and you've, you've seen that before. So right here, once you have it turned on, excuse me, you'll see, the different AI features, that you can set here. So the most granular is what we have set right here is where you have specific users can use specific AI features. You can take a a little step back where just, specific users have all those features you turned on in the previous page or just all users have all AI features that you've turned on in the previous page. Let me zoom back out here a little bit. And then if you scroll down below here, this is where you can set that functionality. So within here, you can either go by again, specific users, a group of users, or even an entire email domain. So depending on what level you wanna set that, you have that functionality there. And then over here you can see I have a group right here. Right now they don't have any AI features turned on, but within here I could say all the features or specific features like a multi select text as you can see here. And really everything that you saw on that previous page you can select. So for this group, I'll just turn all for now. Cool. So that is the control center, just wanted to call that out kind of level set before we jump into some more interesting features here so I'm gonna pop over here. This is just a, an example of vendor bid manager application, that we're gonna use for much of the demo today. Basically the idea of this app is, we're gonna propose, a bid, for various vendors, on certain, products that we might need, for this for this made up business that we have. Right? So in order to access the AI console, you'll now see this little sparkle icon up here at the top. Another important thing to keep in mind before I even go into anything here. The AI chat is gonna contextually know where you are within the platform. So whether you're on the my apps page, whether you're on a dashboard, a report, or a specific record, it's gonna contextually know where you are. So keep in mind you want you may wanna get more specific or less specific depending on where where you're actually at within your, within your realm. Right? It will also, take into account your actual role within the platform as well as in your applications. So if you can build in the app and you can build through the AI functionality in the control center, then you can have that ability but in another app, maybe you can't build, it's not gonna allow you to build there, right? Same deal with like, permissions within data what you can query, what you can't query. Within here, it also gives you, if you come here, it can explore some skills. So this will kind of go through some of those other features you saw in the control center. You know, the ability to actually just create an app from scratch based off some natural language inputs. We give you a few example prompts here, and kind of shuffle through those. Obviously, you can also iterate on that app that you build by creating tables, relationships, formulas, and then you can query the data, create charts, summarize data, get insights from that. You can build pipelines directly through here. And you can also just, you know, ask for help. That's not to say this is all it can do as well. If you can also just come in here and say, hey, what can you do? And then it'll come back and it'll give you a nice little, list of all the different functionality here. Some things I like to do here, is it can actually create a bunch of different types of charts. So I can say what types of charts can you create. You'll notice as it's going through here, it's actually thinking. And you can expand that and it'll give you an idea of kind of how this agent is going about that question you just asked it so that's very, if you're interested on like, you know, depending on what you're looking for, it does its best based off of the context you give it. But just just important to kind of interesting to kind of see what that's doing there. And then you can see here all the different, you know, types of reports and charts it can create for me. Another thing it might not have called out here is, it can actually do some complex SQL queries, between either tables within your application that may not be related to each other or even cross app. So if you want a chart of you know all open issues between several applications, you could do that too. So, pretty cool. But what I'll do right here is I'm just gonna ask it to create me some charts. So I have a little prompt ready to go here. So I'm gonna say create a set of charts leadership would expect to see based on the context of this application. Let's So it's looking at this, getting an idea of the different fields within my app that would make sense. What would be the most critical metrics things like that. Real time demo. We're watching our lives. Yeah. There we go. We got a nice little chart here. You can also see the data in the background there. That'll give you the. It's going too fast for me. Alright, so it came up with quite a bit. Let me scroll back up here. Okay, so this was that pie chart and looking at the data in the background there. You can actually download that data to CSV if you wanted to. We have a few other, you know, it looks like these are the different bids, average bid cost average, like, scores that we have for them. So it came up with quite a bit. Pretty cool there. And then obviously you can continue to kind of vibe against this if you want to get additional detail against this. So another one I was thinking that we would want is based on all the open bids, which one should I prioritize? First and why? You might see I have a spelling error, I probably don't need to make that correction there, but I'm a little OCD so I'm going to do it anyways, it should know that I was trying to say prioritize there. Put it down here. Okay, so it looks like this is a particular bid that it's looking at, then it gives, like, it's giving an urgency based off the date. Because we're approaching that pretty soon. Different statuses that we're looking for. So we still need a few more vendors to get to the the number of vendors that we wanna hit for the number of bids, and then it gives us a nice little recommendation of of how to follow-up. Pretty cool. So the next thing what I'm gonna do, we're gonna go ahead refresh here. I'm gonna open one of these records here. Alrighty, so here's an example of one of those bid records. And I'm just gonna ask it a simple question. So, these are all the different entries. So, basically, all the different vendors that are are proposing what they can do, for us. So we need some lemonade. We need 100, quantities of it, and we need it by, you know, that date. So down below here, these vendors are saying how how much it's gonna cost them and how how many days it's gonna take them to to do that for us. So I'm gonna come back to the the chat here, and I'll just say on this bid, which bid entry should I select and why based on cost and turnaround time. And looks like it's gonna select this one right here, gives a breakdown of why it's the fastest, highest quality feedback score of vendors in the past, Pretty cool. So then, might go ahead and just select that. But right now, what I'll do next is I wanna show a little bit of that building capability, outside of just the analysis that we have here. So as you can see at the top I have this progress bar. So what I'm going to do is just kind of take this existing formula, and just kind of revamp it and make it a little bit more visually pleasing. So I'm going to come back here, let's just create a new chat. So take my progress. Actually, I'm gonna save some time I have a prompt ready that I'm just gonna copy paste for the sake of time here. So I wanna take that progress bar formula, update it so that one, it looks like a modern beautiful enterprise grade software UI of the future. And it goes from a red to green based on the percentage and I kinda list out the percentages that I want for it. So it's trying to find that formula it's identifying the correct formula, if I go into this thinking you can see it actually calls out the correct field ID. It's thinking it's thinking it's thinking some jeopardy music. So, the nice thing here is it's gonna give you that formula output. So, you can review it, make sure it looks good. You could copy this and just create your own formula, or you notice that it gives you that yes or no option so I can say yes and it's actually gonna update that formula for me. And then I can give this a nice little refresh and boom, we have a new progress bar here. And I'm gonna pull up one that's gonna be using a 100% just so you can kinda see the difference let's pull up a report. And then over here, you can see how that updated so before it was just straight green for no matter what it was. Now I can kind of see that it's moving through the different variations there. Pretty cool. And then one last thing I wanna show before I hand this off to Alex. As you can see this is a pretty complex formula came up with right. If I were to say, leave the company, someone takes over, they might not really know what this is doing. So what I can do is tell this to add some comments, to explain exactly what this is doing. So in the progress, our formula, add comments to explain what each element is doing. Some slashes. You can see here it's adding those comments that looks good to me, so I'll click Yes. Give this a refresh. And there you go, it added all of that documentation for me so the next time I come in. I know exactly what this is doing and why. Awesome. So that is the portion of my part of the demo. So I'm gonna stop sharing really quick and hand it off to Alex to do some cool other things with our AI. Nice. Thanks, James. That just just to kinda piggyback off that, been building quick based apps for so long and, the how well that does with edit, like, you know, building formulas and editing your app and making those updates or even recommendations, is just just so so mind blowing. It's it's truly saved myself. And from what I've seen customers do, it's it's really saved them so much time. And, you know, especially when when when you're building an app like this, when you make those edits and you start to, you know, add formulas and all that, like, it can it can really be somewhat time consuming. And so now that, you know, you can just use AI to do that, it's it's really a game changer in in my opinion. Oh, Real quick, Alex. Before you get into your thing, there was one other rich text formula I wanna show off before we get into your stuff. This was something that I just built the, like, the other day. I kinda I vibed with it a little bit longer than than what I just showed you right now. But this one over here, I was pretty proud of. Oh, wow. So this is just like a I wanted like a banner of the data within this record. So this is all all quick base data in the background. Right? And actually even create this little icon for me for each of these, like, this is an app that, like, we're checking different risk areas. So I thought this was super cool. I actually got it to the point where see these these numbers here? These are actually summary fields, and I can click into them, and it's gonna pull up a report of in big yeah. It's gonna pull up a report of those related, records. So that was one I just wanna quickly showcase because I just I'm pretty proud of how that one came out. I thought that was super cool. Yeah. As you should be. Yeah. That one is is is very clean. Probably, that's just one of the the coolest examples I've seen probably so far. Yeah. Thank you, James. For sure. Yeah. Didn't wanna forget about that one. Back to you. No. No. No. Cool. Let me pull my screen here and, kind of piggybacking off of, this this vendor bid management application, to kind of paint the frame a little bit around what I'm gonna focus on is, our AI actions and pipelines and how that can be used for, really, an unlimited amount of of use cases that I've seen, you know, be created in Quickbase. But for this example, you know, we create a bid, and we have a lot of vendors then, you know, submitting bid entries into this bid. Typically, for, let's say, you know, the procurement team that analyzes this and reviews it, that can easily take them anywhere between thirty minutes to an hour to go through, and that's just, you know, per bit. Then you start to multiply that by, let's say, I don't know, 30, you know, 20 bits per week, and that can eat up a lot of time. Right? So in this example, we'll go through some of the capabilities in AI actions and data analyzer. And, yeah, let me just jump right in. So in this case, I have a bid already here pulled up. It's the arcade games. It looks like we already have the two bid entries here. And what I actually wanna do is create a new bid entry, and then I'm gonna pop over to the pipeline to kinda show you what's going on there with this. So in this case, I'm gonna pull up, let's do yeah. Okay. This one pulls up the vendor information here, and I'm gonna say the cost that we're coming in around is, like, $12.07 5. And it's probably gonna take us anywhere between, like, eighteen days to finish. As far as the comments go, I have one already kind of filled out here, so I'm just gonna paste this in. But the the comment now cool. So it's just the typical, process where the the the vendor is now actually submitting their bid. And I'll go ahead and hit save and close. And you can actually then see as well, James, his progress bar has now increased to 75 because we initially requested four, and we have three bids on this. I'm gonna pop over to my pipelining, just kind of explain what I'm doing, and then we'll go back and actually view those results. So, when that new bid entry comes in, we're grabbing all those details. We're looking at the parent bid, that had had gone out to the vendors. We're also then viewing the vendor history, all the vendor history. We're then looking at the competing entries on the same bid. So if there's, let's say, five or six other bids there, we're grabbing all that data, all those child records to the bid. We're then reviewing all that against the new bid that was just submitted. We're then having our AI action step here, the procurement analysis. It's then gonna go through and analyze that new bid against its historical data and against the actual records that are it's it's up against. Right? So the other bid entries that vendors have have submitted. Then we're gonna create a recommendation for that bid entry. Now what's neat about just this kind of a simple workflow is I could, you know, pop in AI. I could ask and say, hey. You know, that recommendation that's gonna be filled out on that record, also send that over to Slack or summarize that even further and shoot that over, you know, in Slack or, depending what that result is, I would like to generate an email and shoot that out to the vendor. There's a lot of possibilities that you can go with this, but just to kind of level set again on the pipeline and the workflow. And I wanna pop back over here to the actual record and hit refresh to kind of show you some of these results. So bid entries, and we'll scroll down here and actually open up the one I had just submitted, which is pending vendor. Let's open this. And now below here, here's my entry decision. Right? And this is where the pipeline's magic really kicked in. So we have the AI recommendation. It's actually saying we need to request for more information with an 80% confidence. That's something that I built into that AI action step. Now I wanna know why. Why do we have to request for more information? We can see here that the bid was around this, which is a $175 higher than the next closest bid. Right? The vendor's historical feedback score is four, which is great, you know, four out of five. But the turnaround time of eighteen days is longer than the event date, which is eighteen days from now. So it actually looks again at all the data, and it actually analyzes that, which is something that the human was doing before. So that's where we're saving a lot of that time, that was previously done. It's also raising some red flags. So the cost, again, it's higher than our competing bids. Turnaround time, that's maybe not so good. And then below, you can see it was generated by the bid entry, triage pipeline that we just, you know, reviewed. So that was just one example. Now we could even take it a step further, as mentioned in kind of the intro here, where, let's say, we're also submitting documents. Right? Maybe some word docs or, you know, a proposal that way. The AI actions would work the same where, okay, it's gonna, take that document, review that document, and then let's say take certain sections of that document and fill in, a quick based record. Right? So your team has insights into that immediately without having to click into a file attachment and actually read through the whole thing. So that's just a quick example of the AI actions. Now to take this use case just another step further, I'll dive into quick insights in data analyzer and how that might be useful for a use case like this. So as an example, I'm gonna pop open my quickbase AI chat here, and it's very easy. You can just click on get quick insights. And for this, we're gonna look at, you know, the bids. We could list all in the field that I wanna focus on. Actually, let's go to the bid entries. Yeah. List all. And I wanna look at the entry status, and the value I wanna predict, around this is awarded. Right? Like, the likelihood that, a certain vendor, would be awarded for these jobs. So I'll let that go ahead and generate. And what the quick insights is gonna do, it's really gonna analyze all that data that I wanted this to target, and it's gonna help me focus in on maybe some insights I wasn't previous you know, there was no knowledge about it before. So it's really gonna kind of bubble up some of those data points. We can see here that the vendor, the premium plush distributors is 75% more likely to be awarded with a very high confidence, and you can see the correlation there. So, again, very neat. We can also see, you know, which vendors, again, are are are really impacted by this and which ones that maybe are better vendors than others. Now if I wanted to take this an even step further with our data analyzer, the difference between the two, one, quick insights is really just doing a kind of a a grab of of the data that you have in your app and showing you more of a high level of, like, hey. Combine, like, this is kind of some of the key insights, again, that we wanna raise here. But if I wanted to get these insights and predictions onto a record level, which then I'd be able to, you know, create the percent fields, the likelihood fields, any other kind of predictions that I would wanna create at a record level, that's where data analyzer comes in. So if I wanted to then, take this into to data analyzer, which I do have open here. I had already created one. Let me zoom in. And you kind of see what I did where it's actually predict the likelihood that an entry status will be awarded. What is the likelihood for me as a vendor, if I was to submit my bid entries, what's the likelihood that I'll get picked and why? Right? So I would wanna dive into that. And if I go back to this record, and I did, again, already add this in, but But I'm gonna scroll over to the right, and then we can see, the award likelihood has then now been populating. I can hit refresh, and and this one should be populating here soon. But we have 46% likelihood, 2020% likelihood. And, again, I can really start to focus in on some of the analytics or the, fields that have, you know, been created from data analyzer to really bubble up some of those insights and to really help drive maybe certain decisions that my business might might be making on the way that we vet and select our vendors moving forward. K. And Alex, the data analyzer created that field for you. Right? It did automatically. Yeah. Yeah. For me. So, you know, let's see. As an example, it did create that. And if I go to new prediction, it does take around, you know, a couple of minutes to actually create because it does have to look through and it actually analyzes all the data, and then it starts to build out this model for you. So I'd actually pick the table, the field, and then the prediction statement, and then what that field should be named. It does all the heavy lifting outside of that for you. So it adds the field to to the correct table, and then you could start to drive those reports from that. And and, Alex, in in general, you need a certain amount of data, right, on here in order to do that reliably. Right? So For for sure. You know, the the more data you have, for sure, the the better. And, you know, I I've for my experience, the better kind of outcomes you see. Because if you only have 10 records here, 10 records there, it's a very light app. There's there's not a whole lot for it to go off of to build its prediction models. Yeah. That's probably great for some of our long existing customers that have apps with tons and tons of data. Right? Absolutely. Yeah. I would highly encourage them to to take a look at it for sure. Yeah. So with that, I'm gonna kind of pivot over to one of the last things I wanted to show, around this because there is really so much that we can do with our AI actions and pipelines. And so I'm I'm open to any questions, you know, anyone has. I'm happy to answer those. But, for the last part here, I wanna pop into our app settings. And Now because as you're making that transition, a big emphasis here is what this can do from you from, like, a revenue acceleration predicting your business. I mean, this is just huge value opportunities. Just wanna make sure I call that out. Obviously, you specifically showed this bid manager. There's obviously a variety of use cases that you'd have you could apply to. And one of the benefits here too is that you can use App Intelligence to understand, like, where could I drive some of this value as you kinda cocreate this with your quick based partner? Absolutely. Yeah. Yep. And and with you know, this this is another great feature that that we've launched recently, which is the app summary. Where this is extremely beneficial, you know, as an example, let's say James is the admin of this application. He's been running it for years. He knows, it in and out all the pipelines, the automations, anything in here that would be important for an admin to know. You know, let's say James is promoted. He goes to a different department, and now I am, you know, looking at this app and I'm like, where where do I start? The app summary really helps answer those questions very quickly. Or if you have a question from your IT team of, like, hey. What what's up with that app? Like, how is it built? What's what data is being entered in that app? The app summary is is great for those type of scenarios where we can see a quick summary that the AI has generated, the key features of the application, the roles and a summary of those roles, and the tables. And, you know, all these summaries, this is not something that is currently in the app. Like, if let's say your summary is blank, the AI will review the schema, the pipelines, the relationships, the fields within the app, and actually then create very good summaries for each one. And I can keep scrolling down, and and I can actually go through all the fields, you know, and as well I can pop into the pipelines. So it's a really great tool. You can export and share, but it is one of those things that has been very helpful just to get grounded on an application if you're not familiar with it. Awesome. Joe, you guys do you have any more questions for me? Yeah. No. It seem like we we don't wanna we don't wanna let let Allegheny anymore. We're gonna open it up for questions, everyone. But, look. Thank you so much for the time. I think you can see there's there's a lot of opportunity here, and this is only gonna get better as our team, you know, produces, or executes on some of these product enhancements, making this more robust and making it so that you have that knowledge there to see even better understand your business context. It's gonna help you, you know, with, achieve some of your goals, whether it's cost savings, revenue acceleration, time management. But hopefully this ROI and this value is really popping out to you. And, you know, appreciate all the time. And, Alex, James, thank you so much. This is this has been perfect. Absolutely. Team, I think I muted myself. Thanks, everybody. The, James, Alex, super effective, helpful demo. Team wanted me to just call out. Empower is coming up, just in, what, a week and a half or so in Houston this year. Please stay tuned and follow us for the keynote. If you're attending, we'd we'd love to see you. I think you can just scan this QR code for for some more information or or just go online. So I wanna make sure we kinda call that out. Big event every year, and this is all about you, right, and giving you some more, context on on our offering. And we'll be covering it was already answered in the chat by, by our team, but we'll be covering the intelligence package in, in several sessions, at Empower as well. Think we're starting to answer a lot of these questions. Alex, James, anything that jumped out to you we didn't get to? I think, Alex, I did, wanna ask you of US context on some of these. Maybe you can just take these in order. Yeah. Sure. Yeah. Trying to figure out where we should start here. I think we answered some questions. I I put some links around data models. That's all online and also around usage and credits. Team, we just try to be super basic on this one and be transparent that everybody's trying to learn AI a bit. We're trying to give you fair usage. So the price you'd pay should be enough to cover almost all of your use cases. And so what we're trying to do is eliminate any hidden charges or fees or surprises and and make sure that really what this is is a, an AI that helps you get more out of the product out of the, Chibi product. Right? So I think that's some of the concept there, but for for extreme details, I'd put a link to you if you could check out online. Then, Alex, I think there was a question or maybe James you could answer. It was about customization of the AI agent. Was that on here? I'm trying to look for it. There's just so many things off to the left. Yep. Yeah. Well, Alex, the you wanna take that one? yeah. Well, the the customization of the AI agent, I mean, there there's a few things that I've personally done so far. You know, one would be, if if, let's say, you're using your AI actions and you wanna customize some of that output, you know, really customize that output whether it's, like, scanning documents and and how it is formatted to then, let's say, read a document and then to put that into a quick base record, that's all customizable on your end. Now as far as the agent in the app, like, we were highlighting a lot throughout that demo, we do have a new feature that's that will be coming out here soon for the knowledge layer, and that, I think will help a lot with some more of that customization around, being able to upload, let's say, internal SOPs or, you know, internal documents and then being able to query the AI agent in your chat console to then, let's say, get an answer, from from those documents that you have in your knowledge layer. And so it is something that that we continue to to develop there. Would love, you know, feedback or ideas on on what you would, want to see, for that customization for that agent. But, those are one thing that we're working on and another way that I've customized those agents in the the pipelines at least. And I think I see a a question here from Matt Chris. Thank you. AI executes a command that, excuse me, impacts raw data, is there a backup, undo, restore feature? So when you when you have a command like that, there will be a prompt that says, like, do you wanna do this? Yes or no? So it's not gonna do it immediately. You will have to have some intervention to say, yes. I wanna do this. But, Alex, right now, I don't think we technically have after you click that yes in undo button. Right? Yeah. No. Not right now. But, I I will say, like, if, a lot of the times, there's almost like a a double gate. So, you know, like, I was just building an app yesterday, and and so I was like, alright. Let let's add some of this data here. It actually, like, creates, like, hey. Does this look right? This is what we're gonna create or these new fields. Yep. That looks right. Then it builds it in the chat first, and then it's like, okay. I built it. Does this look right again? And so there was, like, a two gate, you know, verification where I'm like, yes. Yes. Yes build. And so at that third time, you know, if, you you had those opportunities to actually say, like, yes, no, you know, at that point. And so but I I will say that let's let's let's just say, hey. I added 20 new records as, like, sample data in there after that gate. The both those gates I went through to say yes. It added it, and then I could say, you know what? I actually changed my mind. Those 20 records that were just added, could you please, delete those and verify that they're they're deleted. Right? And so I could actually ask the AI agent to take those away or pull back into what we had before that was executed Yeah. as an example. there's not like a necessarily built in undo, Yeah. but you tell it to undo and it should undo. Yeah. It's funny. Yeah. I even for other done. even other tools that I use, the Alex and Dave, I do the same thing. I'm like, oh, shoot. I hit enter too soon or just say, hey. Forget that last one, and that's some interactivity you can have with it. Yeah. Exactly. Cool. Looks like there's another one. Dataize or get permission before creating the fields. Alex, is the right answer here more just you're only gonna have the permissions that are set to you and the apps you can access? I I about one of your expertise on that one. Yeah. So, I mean, that that field is gonna get created regardless now as as an admin, you know, that you would then go in and and actually see the field in the field settings. Right? And if you wanted to add that field to reports or forms, then you are able to do that. And so that's not something that if you create a date go through the data analyzer, it creates that field. That doesn't mean that that field is gonna be visible to all users. As an admin, you would still go in and actually customize, like, where you wanna place that and and what reports, so you still have that control. Right. And as, an admin only has access to the data analyzer, right, because it's behind the app settings. So you would have to be an admin in the application to begin with in order to to access that that feature. Yeah. Yeah. Or, I mean, again, with the control panel too, you can decide, you know, who who gets access or not. Yeah. I think that's a good call. I do I also wanna reemphasize on the call because I think there's been some cute infusion that I've heard on some customer calls of, like, this control center, you don't have to set permissions on a per user. So if you wanna set something for 50 folks, you can set up group permissions, right, and then just enable it there. I think that's been a common confusion. Like, I don't wanna type 50 names. You don't need to do that. Right? So, James, you provided a good demo today on that, and so you can go rewatch some of that, but I just wanna make sure that it's cleared up and emphasized because that was a huge bit of work that the product team did to alleviate that that a pain point for a lot of you. And then the QBA Excel work really that's our spreadsheet import. Yes. I Yeah. that what you can do is import that, and that, would yeah, the answer to that is yes. So if you have a if. you have an Excel spreadsheet with multiple tabs, maybe you have some VLOOKUPs, it'll actually or whatever kind of formulas, it'll actually recreate those formulas upon import. It'll create the tables. It'll create the relationships. So yes. Yes. That'll work. And I think just maybe the other tangential topic I showed on the slide is that file attachments. The agent can't quite do that today. And so that's something we're working on. The team has a fast file, Excel attachments, PDF, or even images. You've seen other tools that has those capabilities. We know that's important. So that's going to get worked into there and that will allow you to have even more power in the modifications that ends in apps that you create. Totally. And I think. another else is there? yeah. I see another one from Barath. It looks like, does the data analyzer read content from the attached files? To my understanding, no. Right, Alex? But Right. through actions. or the eventual, custom knowledge layer, you could get that data into Quickbase so that it could utilize that information. Right. And then I see another? one from from Meg. Can you demonstrate how QB AI can parse a document and upload data? So we didn't show that today, but that's absolutely something you can do with it. So, essentially, you would have your your trigger whenever that, you know, that file attachment gets added or whatever you want that thing to be. And then the AI action with the, you can set up the the structured AI output. So identifying the different things you wanna take out of that file, and then you give it a a request. So tell it to, you know, extract the data from there. These are the things I'm looking for. And then there's just gonna be an, an option to, like, actually insert the file that you need for it to look through. You'll use the what's called the file transfer handle, within the the previous step with that file attachment. And then if you do it, like, a following update record step, all that AI output would be easily, used to to update the record with with the pills in the the pipeline. Yep. And and I wanna call out here too if, the AI agent is is really, really good at actually helping you and guiding you through that as well. And so if if you were to actually describe kind of what you're looking for for this use case around, hey. I have, you know, this document. When I add it to this file attachment or it's gonna hit my inbox here, here's you know, like, whatever that trigger event is, it will help you build that, entire pipeline. And then there's other features within that building process of, autofill. Right? And so our AI will actually look at the pipeline and start to fill all that in for you. If there is, let's say, an error and it's not running properly, we have an AI troubleshoot as well. And so within that error, you can have AI, like, you know, analyze that, and troubleshoot that, and then it will actually provide the best recommendations on how to fix that. And so that's a very useful tool as you're you're getting started to to build out some of these AI workflows. Let's see what else. Anything else jumping up to you? Oh, I see which ones we answered gonna do a delete. actually. Yeah. I think we've I think we answered this one. It's more just reprompting versus a delete action. Right. The, comment up here at the top go ahead, Alex. oh, I was just someone wanted to see the formula, James, that you had felt. The, which one? The the the banner one? I think so. Yeah. Let's see. It's I able to share end of the at this point? Not sure if I Joe, do we know if I can share through this? I this is my first time using the, backstage, on stage thing. are you gonna do share share the actual are you gonna share the formula? Yeah. If I can. I'm gonna just type it into the question here. I don't, I think you can share your screen. it's a very big blob. Oh, wait. I see a screen over here. Yeah. One second. Yeah. I can pull it up. There's on one lot of buttons. on here. Yeah. I feel like. Nick Fury. share looks good. Sounds like somebody's got a lawnmower in the. background. Yeah. Sorry about that. The landscapers come in, like, the perfect time. Yeah. That's too funny. Hopefully, they're using Quickbase to manage their business. Alright. One second. Almost there, guys. A hallucination rate. I think Alex has a lot of questions going from Carly. Like, you just need to check AI. Right? If if this goes with anything, whether it's Claude or GPT, I I don't really know if we have exact elucidation rate. I think what I would say is that the more specific you can get with some of your prompts and questions, the better output you'll get from there. Mhmm. So it came up with this this big boy and also was able to document it so I know, you know, We we only we? see we see your like, the dashboard with the the actual final formula. No. I don't see. that formula itself. how about now? Yep. I see that. Okay. So this I what it happened was I opened the field setting and it opened a new tab. But, yeah, here we go. So. also, use the the commenting here so I know exactly what what the heck it's doing. But, yeah, this is a big one. This is so cool. Yeah. I don't even know how long it wouldn't would have taken me to to do this with without agent. You go back and forth to test and then go back, iterate, test. Yep. Yeah. Yep. Yeah. So that one was pretty cool. A a cool you know, it's something fun to test. Like, if you already have, an interesting formula, like, an example of what James was showing, you could each just ask a, like, hey. Like, based on this formula, like, how could I potentially improve it to to, like, graphics or to make it look, you know, better? You'd be really surprised on some of those outputs. I had a lot of fun just testing that alone. But, like, how would you like, what can I do to improve this? And then seeing those And I think I see someone asking if that app is available in exchange. No. I actually? just built it in a few weeks ago, know? but, if I I clean it up, I, Well, maybe I'll I'll submit it for you all. there's a I found one. There's actually a bid manager, and it's like a bid management slash project management slightly, and it's it's a little of a bit of a larger app, but it's, it's the same thing. And so that is in the exchange. Jody, I see something. Do you have plans for voice, be able to make outbound calls? I don't believe so. Well, I think maybe the comment I'll get globally on this is that for Fast Field, that team is try is making is is infusing AI across their product just to make the form building and modification and overall use process easier. And so what I will say is they are on the road map putting the ability to speak into your app to record notes, like, when folks are on a daily job site. So that's the kind of context I'll give there, but outbound calls, it's probably a little bit separate, not something we have on our radar right now. Alright. Anything on Quickbase? Alex, James, anything else you see? What about connecting Quickbase apps to other apps to an org, rich validate? Not sure I understand this one. Do any of you have some context? Connecting QuickBooks apps to other apps to enrich validate. Yeah. So, I mean yeah. James, you kinda touched on that too, but that's that is one of the core strengths of of the agent, I believe. It can query, summarize, take action across multiple quick based apps, just using that you know, playing, like, English. So, without requiring users to to do that in order to stitch the data together, that's some yeah. You can. do that. And then, If, you wanna connect of that is, like, say, for example, you had three different apps that have, like, tasks that you need to complete. You could create a chart that has the the x axis being the names of the apps and then the y axis being number of open tasks that you have to complete. And then you'd be able to identify those and go in there, and you wouldn't need to create, like, different connected tables into another app and all of that. You could just just ask the the agent to to create that for you. It looks like oh, thank you. The follow-up there was not other quick based apps, other apps such as websites. Okay. So, Alex, I think that might go back to, like, the knowledge layer and how we could potentially, you know, turn on company knowledge to say something like a SharePoint. Right? Is that I'm thinking that's kind of the direction that question is is heading. Yeah. Potentially. That's what it might that's what jumps out to me because I think some of the things we've heard back feedback on is, like, let's have a bunch of documents in box or Dropbox or SharePoint or something. I'm like, I need this to be trained off of some of that documentation. That's exactly kinda what that knowledge there is to do to enhance the capabilities and the output and context of the agent. So, I think, hopefully, that answers your question, good for you. But, yeah, then that would be the best answer I can give. Alright. I know we're coming up on the last minute. T. Book, I know there's a lot here. We'll try to follow-up on anything that wasn't answered, but hopefully we got the most, if not all of these questions. And the big call out is, we're excited about what's coming. Yeah. They they're this is recording. This is available. We have all this done and through this, through this goal cast tool. So the team, and my team will will make that available. But, yes, thank you. And then, we're looking forward to Empower. So team, please, please attend. We're excited. Thank you everyone, for showing up. Really appreciate your time. y'all. Thank. you. I really appreciate it.