How SaaS companies win with self-service analytics and AI

Broadcasting from our New England studio, Solutions Review is proud to showcase Qrvey in the Solution Spotlight, an inside look at enterprise technologies. I'm Doug Atinson here at Solutions Review and welcome to this solution spotlight featuring Qrvey and focused on how SAS companies win with self-service analytics and AI. AI-driven self-service analytics has evolved from a nice to have to a defining feature of modern B2B SAS products. delivering it in a multi-tenant environment is notoriously difficult, but it doesn't have to be. So today, we will explore why now is the ideal time to elevate your product experience and give your users the power to explore data, uncover insights, and

take action fast. And to walk you through this presentation, we're happy to be joined by Qrvey CTO, David Abramson, for an inside look at how SAS companies are creating valuable differentiation, improving product stickiness, and generating new revenue streams with self-service analytics and AI. David Abramson, thank you very much for being with us. >> Yes, thank you for having me. Excited to be here. Well, uh, I'm glad you're here because at Solutions Review, this is the greatest time ever. Uh, we have technology, uh, hitting in ways that, uh, we haven't seen in a generation. And

I would [gasps] be very curious uh, before we get going actually to learn a little bit more about your journey and kind of what gets you excited about this moment in time. I mean you've been you've been CTO for a while now but you've been in the space I think you started actually probably on the enduser side uh if if I read your LinkedIn profile correctly tell us a little bit more about you know how you've arrived at this at this moment >> yeah no that's uh you're absolutely right I think this is a very exciting time to be building software for sure uh and yeah my background I've I've been either building or helping companies build and deploy software for, you know, the better part of over two decades. And, you know, even here at Qrvey as

CTO, our main focus is helping software SAS companies essentially improve and enhance their software offerings by providing their their end customers, their end users with a comprehensive set of rich embedded analytical capabilities uh that they can essentially ship as part of their own their own products. And you know, in my background, since I've been on the the the build side and helping, you know, uh, deliver software to customers, I know what it's like to have to hit those deadlines and and meet the challenges that that customers are always demanding of us. And so finding ways to do that quickly, efficiently, uh, and in a with compelling outcomes, uh, offers certainly great value. Well, so I know you have a lot you're

going to cover and uh and I want to let you get after it, but I'm going to have a lot of questions, I'm sure. Uh but I don't want to break the flow, so I'm going to turn it over to you. I know you're going to walk through um a few slides and then you're going to get into an actual live demo, which is exciting. So, I'm going to turn it over to you. Uh and when you're all done and ready to wrap it up, just give me uh give me a sign and I will uh pop back in and we'll have a nice Q&A session. >> Yep. Sounds fantastic. So yeah, I'll kick it off. Um, you know, as I mentioned, uh, my name is is David Abramson and I am the CTO here at Qrvey. Uh, and you know, a little bit about my background. Um, I've just been over the last, you know, 20 plus years been working with hundreds of SAS companies to help them get more value, uh, out of

their data uh, by helping them implement, you know, embedded analytics into their own software offerings. And that's really what Qrvey is all about. Um you know we started the business and designed our technology and platform specifically for the SAS use case in mind um to support you know SAS products tenants and users by offering the richest set of multi-tenant embedded analytic features uh that they can essentially white label and integrate into their own uh products and offerings. And you know, we've got a very great collection of companies who are taking our technology and deploying uh and offering these these feature sets and capabilities downstream to their end users and tenants within their their

product sets. Um just to talk a little bit about where we're headed with the discussion today. Uh we've already sort of introduced the topic on why both self-service and AI are incredibly important when building software today. Um, I think the first section we're going to really talk a little bit about why self-service for SAS is really more important than ever. Um, we'll jump into the demo where we actually showcase how you can implement not just self-service but also AI capabilities uh when it comes to your multi-tenant embedded analytics. And then we'll also talk about what that means for monetization and how you can uh essentially generate new revenue streams within your products as well uh with these types of of capabilities within your uh software solutions.

Now as you're probably aware uh being software pretty much everything seems to be trending toward the self-service use case. um whether it is adoption of AI particularly things like generative AI where people just now want to um type things into a prompt and get the answers that way um or even just the overall pressures that product teams are under to help deliver more and more [snorts] value to to their customers. Um the ability to do that in a self-service way offers tons of new opportunity. Um, and we know that the expectations that end users have, you know, are just continuing to go up and up. Um, everybody wants to be able to do things on their own. They don't necessarily

want to have to go and and seek out all of the answers through um, you know, the more traditional means. They want to be able to do things uh, in a fully self-service experience because end users are savvier than ever. They're becoming more specialized. They're more data savvy. um and they're just looking for ways to to leverage software in the best possible uh options for their data. And um this is backed up by you know actual studies and facts. We know that um AI adoption is growing uh as probably everybody uh listening in is aware uh over the last several years especially generative AI adoption has effectively doubled and that continues to trend upwards even though we haven't gotten the actual data

yet for 2025 we know that it continues to rapidly rapidly increase and what that means is SAS companies need to start planning how they're going to incorporate those generative AI features into their tools and products uh particularly uh when it comes to your data and analytics because that's what most people um are using the the generative AI for. How do you how do you get better insights out of the data? How do you create content? Maybe uh publish it and share it and all of that can be a good driver for the types of self-service use cases that customers need. Um so adding those generative AI powered features into your analytic product roadmap. Um, there's a couple of great ways you can do that. Um, whether it's to help your developer

productivity. Um, you know, maybe you're using Jedi to help create the actual content that's going into your products. uh or maybe it's a little bit further down where you're helping both power users and end users either through data exploration AI powered insights or even with uh more chat oriented interfaces to provide things like uh suggestions and recommendations for new uh new insights and new value they can gain from the analytics and the data within the solution. And this really bleeds into certainly self-service because the more AI you introduce, the more flexibility users are going to demand when it comes to self-service within their products and applications. And um as we see here, pretty much everyone now expects that

applications offer your products offer them some form of self-service. And analytics is an excellent place to start because you're never going to be able to keep up with all the demands for the different types of reports and dashboards and visualizations that people need to sort of slice and dice and analyze. Therefore, offering that in a fully self-service way can essentially satisfy these uh customer expectations in meaningful ways. Uh and so what that means and what that ends up looking like in an embedded analytics implementation uh offering those self-service tools whether it's creating dashboards from scratch creating charts and metrics um using no code drag and drop interfaces

creating things like custom filters even maybe several workflows or alerts and the ability to uh collaborate and share with other types of users. All of that can lead to um a very rich and compelling self-service experience within your product particularly around uh the data use cases and as you incorporate more generative AI those use cases can be satisfied uh with these types of self-service tools as well. So how does a company like Qrvey help? Um well there's a lot of things that we can do um to meet those needs and those challenges. As we already know, uh, customers want more. Um, maybe this isn't part of your wheelhouse in terms of your expertise in how to deliver these types of rich capabilities. And

so, uh, companies come to us, they work with Qrvey, and they can leverage our technology to deliver this type of value. And that also means helping you deliver maybe some new subscription options or tiers within your product as well as what we'll talk about a little bit later. How do you monetize these things to generate additional revenue or additional value um from your customers within your software offerings? Because today pretty much every SAS company needs to be uh an analytics company as well. It's not just about collecting data and letting users kind of perform the the day-to-day tasks. It's about how do you create opportunities to understand and gain insights about the data and the types of

things that users are doing every day within your products. Um, and this is it's all about that insight. How do you create um that value? How do you provide a rich set of visualizations, those AI-driven insights, and maybe even some of those automations for customers so that they can dig into the data in the way that they need in a fully self-service way? Um whether it be through, you know, filtering, drill downs, dashboards, and different types of actions or just fully AI powered types of insights that you could potentially offer your customers uh within your product suite. And then from the development standpoint, how do you create more agility um for your product team? Well, you know, product team, they like to

develop, they like to build these capabilities, but how do you get to market fast? Everybody knows, you know, time is money. Uh and then how do you build all of these different types of capabilities? Well, leveraging technology like a Qrvey solution where you can essentially just plug and play um with the built-in tools to provide the self-service and integrated AI capabilities. All of that wrapped with native multi-tenant security within the solution. Uh essentially offers you the fastest time to market when it comes to delivering this type of self-service and AI powered value for your data and analytics down to your customers. And it's all about the growth. Um, how

do you scale that? How do you then provide additional features and services? How do you provide additional roadmap capabilities uh and and those new opportunities to grow your revenue and grow your product suite um for your customers. So all of that can be enabled and powered by the Qrvey platform and technology embedded into your uh solution. And last but not least, we can talk about the AI feature sets and AI capabilities um that really allow you to get to the most compelling end-user experience within your product. Um it's all about being able to deliver AI-driven analytical results. um being able to you know find information with you know

natural language being able to integrate with APIs and all of the data and security models that you want to have and then how do you implement all the different interactions whether it's you know through tool sets like you know LLMs and and the models and prompts that those will be driven by or um more developer oriented integrations like MCP servers for instruction sets and how do you power um your product through a different set of of AI tools and capabilities within your software offerings. Um so all that sort of blends together in creating that next level of self-service interactions that you can offer downstream um to your customers. Okay. So, at this point, we're going to take a look at, you know, some of

examples of how this could uh be displayed and be rendered, embedded into your um product suite with a quick demo of some of the interactive self-service capabilities uh within the Qrvey offering. So, I'm going to jump in real quickly into a couple different scenarios here um for how Qrvey can work uh when it comes to incorporating both self-service and AI powered capabilities in your application interfaces. So, what you're seeing on screen right now um this is a pretty typical interactive embedded data visualization dashboard widget control powered by the Qrvey stack. Um, as you can see, we have a lot of metrics that we're looking at. We

have a variety of different types of chart visualizations and everything here, um, supports, you know, different types of interactions, whether it's, you know, click actions to filter or leveraging even custom visualizations to drill and filter on the data or drill down into different levels of detail as users are exploring the data and information. So, tons of flexibility to offer users a variety of different options for them to select um integrate with different uh data properties and adjust the data information that they're seeing. Um these are all kind of very basic types of interactions that we're leveraging where whether it's a filter or a drill down maybe you want to explore the data with a sort of a drill

through um into underlying uh information or details um that you're looking at within the interface. So tons of flexibility to kind of offer a different sets of of interactions and details u within the UI. But in addition to this, when it comes to what we call self-service, it's all about letting the users customize their own experience. And what does self-service customization end up looking like? Well, you can do lots of different things. You might choose to let customers change the data that they're looking at. So, instead of grouping um by one series, you could group by a completely different series. Instead of looking at data in weeks, you might let users change the aggregations

and how the data is uh calculated. And so tons of flexibility to offer different options um on the fly within the within the interface. This could also mean even changing the type of visual that we're looking at. So instead of uh you know the the columns chart, we could adjust it to look at maybe a line chart presentation or really any other type of presentation that might provide a little bit more meaningful access to the data and the output um that we're working with. This could also include you even doing things like adjusting how the data is sorted um and changing even the types of information that we're seeing um you know within the the different axes or the different properties of the detailed data that

that we're working with uh within the UI. So a lot of flexibility to let users kind of customize adjust but this is all from an existing template. You know what? If we said, "Hey, I don't want to pre-build any of these types of controls for my customers. I want to let them drive their outcomes." Well, that's another layer of self-service and features that you can offer to your customers with embedded Qrvey technology. Um, so instead of just making those quick adjustments and changes, we could start from this template, but we could also create from scratch. And that's where we get into the true power of self-service where letting users actually design their own

components, collaborate with others, and even publish those out. So other colleagues or other members of that same uh customer or tenant group can actually share and see um the types of things that users are building within their organization. So here we have again kind of a blank canvas as a starting point. Uh we can either inject maybe you have a library of existing chart controls um that you want to select from and and I can pick and choose um from this existing library and I want to just load these into the canvas and start sort of organizing and drag and drop and creating the components or controls um that build those reports and dashboards for me. Um you could build charts from

scratch as well. Create a new chart. I select the data that I want to work with. And again through uh basic drag and drop techniques. I can select the data that I want. I can choose uh the different value series. I can aggregate how I want to aggregate the data. Uh and then from there I can style it and customize it in any way or I could even do things like adjust or change the type of visualizations that I want to work with. So, the full suite and library of self-service tools are available to me to be able to generate and customize and create any of the content that I want to see. But now, as we get into the age of AI, what if my users don't want to have to use the same drag and drop techniques and have to understand everything that's

happening sort of within the data model uh before they start building any of the outputs. Well, another option that we do provide is to use generative AI prompts to create any of the output and content. So, instead of dragging and and selecting the data that I want, all I need to do is point it at the data that I want to analyze and using the sort of open-ended natural language prompt, I can just ask questions uh to gain those insights. So maybe I want to learn, you know, what is my monthly trend um based on the data that I'm working with here. Um and based off of this, I can simply uh

enter that query into the prompt and let the system sort of do its thing. Right? We're pointing at the um large language model. Um, and based off of that, it's going to return um the output that matches um the typed in prompt query. And so essentially what it's doing, it's explaining how it got to this outcome. It's saying to to analyze that trend, we're going to look at this line chart. We're going to use the dates and group them into those months. Um and then we're going to count up um the details and uh essentially add that series um so that we can uh differentiate across the locations within our data. So just like that, you know, I I enter the prompt, it generates the output that I want to see and if I'm

satisfied with this, I can easily add that into um the visualization output and share that uh with other users or uh with any sort of generative AI tool. I've created a conversation so I can continue to iterate, ask additional questions, um, change things about the visualization, um, and then use the prompt to continue to iterate on top of of the outputs that we're working with. But maybe I'm satisfied with that presentation. And then you still have the ability in the full self-service mode to go into the more traditional editing style models. Uh, maybe instead of months, I want to look at weeks. uh maybe I want to change how the data is summarized or maybe I want to look at a different metric um that we're working off of. So being able to iterate and

make those changes gives me a ton of flexibility um to kind of adjust even once um you start autogenerating these components through um the different AI services uh that you're working with as well. Now on the other side, we can also use AI tools to further explore the data once we've already generated um the outputs in our visual presentations, our reports and dashboards. And so it's not just about using AI to actually build and create the content in the full self-service model. um once we already have our outputs maybe I'm interested in understanding a little bit more detail about what's happening within any of these these visualizations and so from

that I can actually analyze these visuals using AI services and in this model what the AI can do is it can even suggest some insightful questions I might choose to ask so um you know maybe I want to leverage some of the questions that it's uh sort of predefining for me and could say, "Hey, you know, how do I look into this data in a little bit more insights?" Or maybe I just want to use the prompt to ask other questions um that the AI system hasn't thought of. And so this is now how we can dig into the data in much more detail um using, you know, the different AI services incorporated into our enduser applications. Um you know, so maybe I want to look and see, you know, why is

there a spike in uh Q2, right? So, I want to understand a little bit more about um how that's going to work. And so, I can allow the AI generative system to help me analyze that data in a little bit more detail and then give me some insights into why I might have um you know, some of those outcomes um that I'm I'm trying to understand. So, in this case, you know, maybe it was due to some of the seasonality or maybe some promotions that I was running. And then because of that, I can then take that and continue to ask further questions or I can use that to inform the additional self-service outcomes that I'm going to want to generate within uh my additional uh charts and outputs in the UI. So, as we can see, there's a lot of great

opportunities to introduce not just full self-service style capabilities, whether it's designing uh interacting and creating content to leveraging those integrated AI based services, prompts to help generate content, prompts to help dig in and analyze content in much more defined details, and all of these come together to create that additional value um for your customers. And as we kind of wrap up the demo here, um this is exactly why SAS companies choose to leverage technology like what we offer with Qrvey. It's all about that embedded fully customizable self-service set of capabilities to

let your customers create design not just sort of the dashboard outputs but integrated with the AI tools and those type of capabilities. Um, and then also leverage that with the full stack of technology, the data engine, the full secure, scalable multi-tenant architecture, and all of it can sit and and live side by side with your product deployed natively into your own uh cloud uh infrastructure. So um kind of get the best of all worlds when it comes to not just the enduser interactions but the scalable technology stack and um the fully deployed solution that lives and and uh operates entirely within your uh

tech stack as part of your overall um product suite and product offering. So last thing I just wanted to kind of touch on as well is now that you understand a sort of what's driving this innovation and and why you need to be planning for these new feature sets and capabilities as part of your software offerings because of you know the AI capabilities that are constantly evolving and innovating plus the demand for self-service and then we saw how that can be implemented with different self-service capabilities in the demo of some of the capabilities inside the Qrvey platform. But then how do we introduce that into our product uh and even help monetize that to create additional value um for our customers?

Well, we have customers that have already done this and so let's talk a little bit about some of the models and some of the strategies um that they went through um to showcase how that can work within a typical SAS solution. So, this first example that I'm going to sort of run through here um is all about how to create um models for a value-based pricing increase. And for this customer, their goal was to essentially differentiate their product obviously with the self-service and um AI generated capabilities uh but then um use that to charge for additional add-ons above the sort of base tier of

the subscription into their product. Um and overall their goal was um to essentially hit um some pretty aggressive additional revenue coming after that uh first year of initial product development investment and the way that they differentiated their product was was in three tiers. So the first tier was um basically the the pre-built components. So accessing all the historical data and generating pre-built dashboards and embedding those into their software solution. Um the second tier was introducing more data types including even some external sources of data um along with offering

customers real time data access and real-time analysis of the data. And then the third tier finally introduced the ability to monetize the user created content. So self-service analyses, how they're building their own content and then allowing those users to create essentially unlimited outputs. So um the ability for users to just get in there and do whatever they needed uh to construct their own sort of self-service analyses as part of the product offering. So those three tiers really helped them to differentiate and create new opportunities to upsell and charge for those add-ons when it came to delivering the software uh capabilities in their product.

Um the second example here was a customer who looked at it based more on a volume uh pricing model. So, um, instead of just the value ad, uh, tiers, um, it was really about, um, factoring in all of the the total volume of not just the usage, but the the overall customer license, um, that they're working with and then extending that um, by offering these additional packages of tools and solutions. So the base package was just access to um some of the essential self-service um for users to edit um view, export, and create, modify. Um but they limited how much uh actual content could be

created. So imposing some restrictions on what users could do. Um for users who wanted unlimited access and unlimited create opportunities, um that was the enterprise package. So they could create kind of uh again a volume based model uh with limits and no limits on what the users could actually do within the solution and then they certainly leave themselves open for uh future add-ons. So adding capabilities like automation and workflows, additional tools around forecasting and detection of outliers and then certainly all of the AI suite um uh features. So the analysis uh through AI and the design and creation of components um through AI as well as

all of the backend integrations with uh APIs uh and MCP services um that are available in the Qrvey stack um as well. So, lots of opportunities not just to create the initial packaging models but then also adding future uh growth opportunities with additional add-on tools and additional packages that could come uh with the extended roadmap. Now, um typical path to how that monetization strategy can work. Um these are just some sort of rough timelines for what it looks like as you're planning your next steps and initiative. You know, oftentimes it's going to start with a proof of concept. Um, and then we kind of get into the meat of the implementation from um the onboarding,

which is all about mapping and planning the solution to the MVP launch and maybe a limited production rollout to the full expansion into uh a complete self-service implementation across the entire uh user base and tenant workspace. And uh these are pretty typical timelines for what we see um our partners uh and and their uh go-to market strategy and how they're getting uh into production in front of their customers uh leveraging the Qrvey technology as part of their embedded analytic offerings. Okay, so let's talk about where that leaves us. Um obviously um you know we've seen how much AI and self-service

can be changing the way that you need to be thinking about how you're going to deliver value to your customers. Uh and we hope that you consider companies like Qrvey um to help you get there. Um, you know, we like to think about it as future-proofing your software because again, um, oftentimes, um, you know, you can kind of get bogged down in answering customer requests and not necessarily thinking about what are all of these exciting new opportunities, particularly in the world of data and analytics, generative AI, and self-service. Uh and so the robustness of a platform like Qrvey and our continuously evolving roadmap helps you create um essentially a integrated software strategy um to not only deliver

value today but also to continue to deliver value um as you iterate and and release new versions of your software. Uh and then thinking about how you can use the technology to help create new value uh better monetization opportunities within your software. Um and as you saw from you know the previous slide um you can get there often times much much faster than you think. Uh because a lot of the stuff can be very easily integrated and dropped right into your uh production solution in a matter of you know oftentimes weeks. Um, so if you're interested to learn more, um, I'd certainly encourage you to, um, reach out. Um, you can certainly book a demo. Um, and one of the things

that we're also providing everybody with is a comprehensive embedded analytics evaluation guide that you can use as you start down this path of figuring out how you need to evolve your software and your technology. Uh but we do also have a ton of other tools uh including a developer playground um our interactive vendor score card and certainly um some additional tools around uh ROI and how you can measure that a little bit easier as you're uh beginning this evaluation path to understanding how embedded analytics can help you enhance and improve um your overall um software offering. Um so that's it for my slides and presentation. Um, but you know, I'm I'm happy here to answer questions or

continue the discussion with whatever direction you want. >> Well, I I I actually but there's a lot that I want to dig into, but I don't want to skip over um a point that I made in the intro and also a point you made in your intro, which is this idea of multi-tenant analytics. Why is that hard to do? Because I believe it is hard to do and it's kind of a I don't want to bury the lead there. That's that's kind of a an important element to all this. Correct. Oh, absolutely. Um, yeah, m multi-tenant is sort of the bread and butter of all SAS, right? So, um, in almost all cases, your SAS application is going to be serving multiple tenants or multiple customers. Now the reason why multi-tenant analytics is difficult is because traditionally

analytics solutions. So if you're looking at commercial off-the-shelf products, they're all designed for um enterprise or internal analysis of data. So think uh a single organization who wants to gain insights about how their data is performing or how their business is performing. Um they're using these technologies, these products to help do that. Um however the SAS implementation of analytics was sort of overlooked in that use case right um you have kind of enterprise analytics which is great for an individual organization or individual company but it doesn't really match the needs of a SAS business when it comes to analyzing data for multiple companies. Um, and ultimately maybe you even want

to do things like benchmark across industries or benchmark across geographies or do these things where you're taking data and not just using it to analyze for the betterment of a specific group or a specific tenant but also using it to sort of uh match up or or benchmark yourselves against um the the industry or some other segment of of people. And you have to do that in a way where the data remains secure. uh and you have to do that in a way where access is guaranteed that you're only accessing the information that your company or your um your role is allowed to access uh within the solution. So um that's why it's hard. There's security challenges that people don't think about. There's data challenges that

create a whole new world of complexity when it comes to multi-tenant implementation. Um and you know these are all the things that uh we've thought about. So at Qrvey um we started from that goal in mind to create a comprehensive solution around supporting a multi-tenant embedded analytics use case and so everything that we build in the software is designed for that specific need. So so security um ro and access credentials entitlements all of that logic is is configured and can be configured to support a multi-tenant SAS implementation. So, it feels like um we're in um middle earth with regard to where we're headed

that that self-service uh and uh low code, no code, all the all these terms that were kind of the last uh five or 10 years are now morphing into agentic AI for all intents and purposes. And and I think the end state that people are kind of predicting is is kind of autonomous agents that begin to operate as data analysts. I don't know where do you see um where do you see this all kind of in the current moment and and and what are going to be some of the limitations uh with regard to trying to get to that end state and do we really ever get to

that? Are the agents really coming? >> Uh, so they're definitely coming, but I I think it's important to caveat it with um it's really still all about the application, right? So without the application, the agents can't really do anything. And I think that's really important to understand about how AI fits into certainly software and software development and end user experience. uh because if they don't have the right instructions and the right sort of outcomes to seek out um you really don't have a lot of utility for many of the the AI tools and prompts um that could exist. And so I think it's all about mapping the AI services to the right application use cases. And this is where

I think certainly SAS businesses, SAS companies have an amazing opportunity to provide that value because if you can orchestrate your products through AI agents, prompts, general AI services, that's a fantastic opportunity to sort of alleviate the um burden of self-service from the coder. Right? So now all of a sudden you don't have to create all of these new interfaces and drag and drop tools and all of these ways for somebody to navigate a UI. Um you can essentially just use the prompt as the interface and have it talk to the outcomes that your application is

expecting. And I think that's the amazing opportunity that exists today for any SAS company is to start incorporating that as the interface for their products. And because you already have the outcomes that people are expecting, whether it's an application that's doing inventory or if it's an application that's doing um you know CRM style customer management or maybe it's a in a healthcare setting. All of these applications have very well-defined outcomes that traditionally you would have to build interfaces to be able to to leverage. Now with generative AI and all these more promptbased and agent-based models, you don't necessarily have to spend a huge amount of time constructing new interfaces, the

interface is the prompt and your application and the work that the application can do is sort of the outcome that the prompt can help you achieve. And I think that's a fantastic opportunity immediately. It's not this isn't the future. This is happening now. And so where does it go in the future? I think you have a lot of opportunities now to then continue to extend what your application can do because you don't have to worry so much about building all these new interfaces. Um you can worry more about what are the additional outcomes you want your application to have. And that's why I think the data side of it is really compelling because you have all this data already. You don't need need to necessarily build all these new interfaces. you can just essentially point the AI at the data and let it uncover the insights for the customer for the users or even for the

development team to help them build more application outcomes. So yeah, there's some really, you know, interesting opportunities that exist immediately and then all the future opportunities that can come uh once you start essentially uh automating and and I don't know AI all the interfaces of the of your systems. Well, I mean, you showed it uh in one of your slides, the adoption of of generative AI and you know, how human behavior ultimately kind of seeks uh ease and comfort and so the promptbased uh interfaces and promptbased uh applications, tools uh enablement

uh is going to be table stakes. I mean, there's there's just no question about it that the more people use generative AI, the more they're going to expect to be able to use a similar interface within their other software applications and expect it to behave in a similar manner by prompting other prompts, I guess, is a is a is a way to think about it as you demonstrated. Um, but I'm curious about how that integrates. So, so the other the flip side to all that and uh and you mentioned a little bit about security, but certainly the flip side to all of that is where how does the integration work um where are the kind of limitations um in terms of what it can produce in terms of prompting with the access to the data that it has?

How much does it seek the LLM data? I mean where where where where have you found the kind of [snorts] you know going as far as you can without necessarily compromising on the on the security? >> Well, security is certainly you know one of the most important aspects of any sort of um SAS software solution and in the development mindset of how you need to make sure that you know obviously users are only getting access to what they're supposed to have access to. Um, you know, with that in mind, I think one of the things that we've done at Qrvey is the way we integrate is a fully deployed solution. So, even though all of our customers operate as SAS businesses, um, Qrvey because we're so close to the data, um, we don't operate as a SAS, we're actually a fully

deployed solution, uh, where it lives essentially alongside your own company software. Um and so with that the security is is essentially inherent to deploy to the deployment model where um you don't have to send data outside of your firewall. You don't have to send data to third parties. Um it can all live together. And I think um it's an important thing to understand where the data is going when you're doing anything with these um you know interactions and uh feature sets and capabilities. And certainly the AI implementations are no different. Um again that's why uh even the way Qrvey has implemented you know all the AI services it's essentially a bring your own model style model um where you can point it at your existing

um AI models or if you want to use you know any of the publicly available ones that's certainly fine um to leverage uh you know you you want to leverage open AI's options or any of the um you know Google or AWS or Azure services all of those are um available to you to leverage as part of the the integrations. Uh but it's key. Yeah, security is very very important in this world and um I think the opportunity that that we're offering is to maintain all of your own data privacy and security because of the fact that it is a fully deployed solution uh within your own um application firewalls. Well, and so there is a uh there is a bonus gift for all the attendees uh with regard to

receiving Qrvey's comprehensive embedded analytics evaluation guide for SAS leaders which we'll have up on the screen there uh and you will all receive for being a registered uh attendee. Um but I wanted to ask about the implementation timelines around um self-service analytics. So, so how does it, you know, after after people have seen this and they're jazzed up and they they come through the front door and and uh and engage, what does it look like uh on the implement implementation timeline? >> Yeah. So, I did share some rough kind of timelines that we've seen across our implementations with our current partners and customers. I I'd say it's

really a uh essentially a tiered or sort of a phased approach typically to to anything you're going to do when it comes to embedded analytics with the first phase being how do you just validate that you're going to get to the outcome that you want through something like a P. Um, usually the PC is also something that can be used for consensus building, whether it be internal team consensus, say, "Hey, yes, this is what we want to we want to continue to work on and build or maybe it's how do I get this in front of some customers to pilot it and say, hey, is this what you guys are looking for?" And then we can continue to iterate on top of that. And um, that process is typically something that takes no more than two to three weeks. you know, we've seen uh especially with the with the Qrvey model, a PC um

usually, especially if you're you're you're using it for that sort of consensus building model can be wrapped up in a very very short time frame. Um and then once you kind of have the end goal in sight, uh it's all about how do you do the implementation? And because we've created all these great tools for not just sort of building and planning but also publishing uh of the interface and uh the ease of integration, you know, we're really talking about um weeks instead of months um to even get the uh initial product launch and uh implemented. So, you know, we've had some customers get to market in production um with a full set of end userfacing self-service tools in in a matter of about six weeks. Um so it can

be very fast uh turnaround once you kind of know and and understand what you're trying to achieve. Uh you can get there pretty quickly and then uh from there kind of the iteration is is the sky is the limit, right? You can you can take it as far as you want or you can continue to kind of launch new new versions and capabilities and uh decide how you want to package this up and and certainly monetize those if you need to in the future. So, um, it's interesting our perspective, uh, from all the things that we're seeing is that there's a, it feels like there's a a a decent segment of the enduser market that is almost paralyzed by this moment where they're afraid to kind of purchase um, anything while so much change is

happening and there's rumors and investment about all these other things that are coming and and I'm curious about your perspective because you're seeing things being built um and you're seeing products succeed more than others. Um what do you what do you see the market um gravitating toward? I mean what are what are some some of the larger [snorts] you know functionalities I guess that that uh that are gaining traction with some of the SAS companies that you're partnering with? Yeah, I mean it's certainly interesting time right now because there's a lot more scrutiny on certainly um every aspect of the business whether it's you

know just overall you know customer satisfaction to customer retention to how do you grow uh but that is also being felt in developer organizations I would say you know a few years ago what people chose to build and how they went about building it was probably far less scrutinized that it than it is today uh within any development organization. And so you have to be able to justify what you're working on more so than ever. And I think we're seeing that becoming easier and easier once you start looking at these opportunities that do exist with both what I you know what we are defining as self-service as well as how you can incorporate

different types of AI components or AI prompts into your application interfaces because once you start doing that you stop sort of chasing your tail in the sense that you know I I've seen companies really just kind of get bogged down in customer requests where they say, "Hey, customers keep demanding these things and now I need to just continuously iterate and respond to what the customers want want from me." Where you can sort of shift the burden, right? You can say, "Hey, instead of continuously working and playing catchup to what customers are asking for, how about I give them interfaces where they can essentially do it themselves?" Um, but I don't want to give them work. So the more interfaces I can create that are easy, especially through things like

AI-driven prompts, you know, the faster not only I can give them what they need, but the easier it is for them to continue to iterate without having to, you know, bog my development team down with all these uh requests that kept, you know, coming in. So um it helps not just justify your your end goal, which is to give your customers higher value, but it also helps you alleviate the burden on your your engineering teams. uh because now you're not continuously just responding to the requests that are coming in from the field. >> Yeah, it's it's an it's an exciting time. Uh and this has been incredibly uh interesting and beneficial. And again, we would encourage everybody who's registered for this to engage with Qrvey and and and have these value added conversations. You're not only going to

be able to talk to one of the leading solution providers uh that can handle this, but you're going to get to get some more perspective from uh from all the other things that they're working on and seeing. Uh and you can't sit still. This is just no time to be sitting still. You have to you have to start building. You have to build better. And uh and so best of luck, David. This is uh like I said, been very informative. Uh, thanks for coming on. Uh, and I can only imagine what the next couple of years are going to be like, but uh, but get some sleep as well. >> Fantastic. Yep. Thanks for having me. >> You bet. So, there you have it. Another solution in our spotlight. We want to thank David

Abramson and Qrvey for that presentation today, and we appreciate your attendance as well. Until next time, I'm Doug Atinson here at Solutions Review. Thanks for watching. If your business would like to be featured in a future event, contact us today.