SAS Innovate 2024 | Jared Peterson | SAS

Clips
More from SAS Innovate 2024

Jared Peterson

SVP, RND

SAS

play_circle_outline Exploring the Power of SAS Viya Workbench: Harnessing the Blend of Open Source and SAS Analytics

play_circle_outline Native Python libraries and SAS analytics in Workbench

play_circle_outline Lean and fast user experience

play_circle_outline Viya Workbench: Enhancing Data Exploration with Self-Service Architecture, Cloud Scalability, and Open Source Integration

Info

Jared Peterson | SAS

Jared Peterson

SVP, RND SAS

search

>> Welcome back, everyone, to theCUBE's live coverage here at SAS Innovate 2024. I'm John Furrier, host of theCUBE. My co-founder and co-host, Dave Vellante, extracting the signal from the noise, report on all the top stories. Go to siliconangle.com, check out all the stories. Of course, thecube.net for all the viewers, as we're in day two of two days of coverage. We got Jared Peterson, senior VP of platform engineering at SAS. He was on stage where he did the demo and the announcement of the general availability of SAS Viya Workbench. Welcome to Back to happy to be here. You'd be hard-pressed to find something I'm more excited to talk about than Viya Workbench.

>> One of thing I love about the product, and just where it's going, is it's really teed up perfectly for both the AI wave and developer needs. If you look at the market, and we report this on SiliconANGLE and cover it on theCube with other experts, the rise of open source in the AI world on the model side is merging and almost connecting with the proprietary models from size and functionality. And also the requirements to run them are getting easier, you're seeing some models running on an iPhone.

Jared Peterson

>> Sure.

>> We just talked to Intel about cost of ownership. So the developer community is booming.

Jared Peterson

>> That's right.

>> So having a Workbench is a pretty big deal.

Jared Peterson

>> Yeah, Workbench. You talk about the developer community and the rise of open source, and you can really see Workbench as us trying to reach back with our deep legacy in these areas, but then bring some of that forward and say, Okay, can we bring a tool to the market that blends both the world of SAS, SAS programming, but also if you want to just do pure open source, or you want to use open source languages like Python, but use SAS underneath the covers. Workbench just brings this all together in this amazing package. And we're really proud of it just because of all the things

>> What's the big highlight? What's the big table? People watching right now, they're SAS customers, or a prospect. First of all, analytics products well-known, and decades of success. What's the big deal? What's the high-order bit of the news?

Jared Peterson

>> Yeah. I would go through a couple things. First, I think about just Workbench is this cloud native, cloud scalable, rapidly accessible development environment. Then, once you have a development environment open, it brings the best of breed tools from the open source community to your palette, so to speak. So you've got Visual Studio Code if you want to go that angle. You've got Jupiter Lab if you want to go that angle. And so, we're bringing those IDEs to the table. Cloud scalable, cloud native, got these IDEs, but then we're bringing SAS analytics into all of that. If you're a SAS coder, the SAS language works beautifully in both of those editors. If you're a pure open source developer, Python is there underneath the covers ready to go right when you start up, you can use it in both of those editors as well. But, then if you want to say, okay, I'm interested in the SAS analytics, the reliable analytics that SAS has been known for for years, but I want that in Python. We've now got native Python libraries that are also there in Workbench. And then the last thing I would say for the SAS coder is, in Workbench we've taken all of these analytical investments we've made in the last few years in Viya that are in what we call our CAS server, via what we call actions. Things like advanced computer vision algorithms, advanced natural language processing, neural networks, those kind of things, deep learning. And now, all of that's being exposed in the form of prox. And prox are the unit of value in the SAS language. So all that capability that's been in Viya is now in the SAS language. So SAS 9 coder can come into Workbench. They know the SAS language, they get all this new goodness, and it's all there in this environment that's just lightening fast and starts up Workbench fills a nice hole in the portfolio.

>> And I'm interested in how it can be used as a lever to add incremental value to the roadmap. What can we expect now that that hole is filled? And what kind of flywheel effect can you get in the roadmap?

Jared Peterson

>> Yeah. When we look at the roadmap and you see us doing some things in Workbench that are absolutely hints as the things that are coming in the future. One of the things we like to think about in Workbench is it's like this leanest, meanest rendering of the Viya platform as possible. We've taken out anything and everything that you don't absolutely need for this pure developer persona, this pure SAS and Python coder. And so we've just leaned it out. And so you'll see us do more of that we have brought people to Workbench. They're then going to expect that experience of, man, this thing starts up fast. It's blazing fast when I'm executing code. The way we got that experience is because we leaned all those things out. So now let's say, okay, maybe they want some visualization capabilities on top of that. Maybe they want model management on top of that. Well, we've got to make sure that we also render those things out in the same way we've done Workbench so that you get all that the same. It starts up in seconds, it runs super fast. And so, yeah, it is that bridge. It's like, okay, let's bring you in. Let's get you connected to all these things, all this goodness. And then, let's bring the rest of Viya to the table.

>> And you got to keep it lightweight though.

Jared Peterson

>> That's right.

>> So architecturally, how do you do that?

Jared Peterson

>> Well, the way you do that is you only start up, you only bring to the table, and from a compute and resource perspective, the things that the user absolutely wants. The things that they're demanding. Things aren't just running all the time in the background. And so maybe they start with code, the pure Workbench experience now. And maybe they're using code for ETL, they're getting data in, they're transforming that data via SAS or Python. But then at some point, maybe they're ready to do some exploratory analysis of that data. And so at that point in the future, it's like, okay, as soon as I have that thought, I want to do this. That's when it's like, okay, let's now just go spin up the containers behind the scenes that are necessary for that experience, and only that experience, so that then get that same snappiness and performance.

>> What are some of the customer early adopters saying about the product? Obviously, I liked the UI. You guys nailed that. Great job on that. It looked easy to use. I haven't used it, but demos you guys gave, I was blown away by that. So user interface, check. I'm assuming they like it.

Jared Peterson

>> Yeah.

>> What's some of the other feedback?

Jared Peterson

>> Well, on people like it question, the first thing to note is it happens over and over again, both internally at SAS, but then externally as we interface with customers. The first time people interact with Workbench, before we even get to feature function, they just smile. It's just this delightful experience. And so that's, as a developer inside of an RD organization, that's always just fun. And you've got something when you just see this joy wash over people. But then beyond that, the thing that we hear people tell us is, you see this light bulb come on of, Wait a second, in the same platform I can enable SAS code development. I can enable pure open source development. I can enable open source languages using SAS math, and in the future other additional languages like R.When they see that they can do that, all that in one tool, that's when they're like, Oh, wow, okay, this is different.so that's the biggest thing we hear is people realizing they can bring all that underneath one offering. And so when that light bulb comes on, people get you said R is coming, yes?

Jared Peterson

>> Yes, absolutely. Later this year, yeah.

>> One of the things we see is you guys always had been talking about democratization.

Jared Peterson

>> Right.

>> Now you have a real developer product. And so I was intrigued by the slide that said, Tool for the data science developer.The data word data science developer. Now, in this world of the population needed for data scientists doesn't need to be big.

Jared Peterson

>> That's right.

>> I won't say shrinking, but you don't need to have these alpha nerd geeky data scientists jamming all the time as the table stakes. You can hire one or even use a managed service. Now you have the more user base coming on, which opens up to the developer. Classic app developer.

Jared Peterson

>> Right. Right. Yeah. No, exactly. Generative AI coming into tools like Workbench, whether it's through co-pilots and code assistants and those things is going to enable people maybe coming from more of just a developer background to then dabble in data science. So that's one way to go. And it'll be interesting to see how generative AI across the board allows developers, data scientists to level up in whatever dimension they're interested in.

>> Jared, one of the things that came up in our analyst roundtable this afternoon was, the recognition that the old days, the tech genius builds an algorithm. They company goes public. They have core technology. New technology is ubiquitous, it's now the glue that holds things together, and everything around is the new IP. And of course, we've been saying on theCUBE workflows and data new intellectual property. Which is tied into your theme. But now when you start opening up the democratization to developers to code with Workbench, the rock star, the superstar, the person could be business user.

Jared Peterson

>> Absolutely.

>> It's like, now it's shifting to not getting a little anything for the team award or a little pat on the back, or a little bump in promotion or raise. Someone in the org could be a game changer on the business model for the company.

Jared Peterson

>> Oh, for sure.a huge power dynamic.

Jared Peterson

>> Absolutely. One of things I like to remind people of internally at SAS is, whether it's with tools like Workbench or just what you're seeing happen across the space with generative AI, is that, yeah, it doesn't have to be a developer, it might be a business user. But I like to remind people, in the future we will be competing with companies we don't even know exist today because these tools enable this wave of productivity that we've just never seen and never imagined. And so, teams of one, teams of two, teams of three are going to be able to do things that we would never thought possible years ago.

>> And I the intelligence being built and devised, another one I like. Another announcement that... First the models, selling lightweight models. That's a home run, by the way. I think that's going to be excited about it.

>> The prompt saving concept is interesting, because that is a nuanced point, but it brings out the direction of saving stuff and using it later.

Jared Peterson

>> That's right.

>> And prompt-less potentially around the corner, a prompt-less environment. It just does it for me.

Jared Peterson

>> Right, right. Yeah, no. We're really excited about the prompt catalog work. And one of the reasons we're so excited about it is it allows us to blend both new generative AI, which of course is dependent on prompts, that's the way you interact with that system. But a prompt catalog allows you to bring what you might consider to be more traditional NLP task and technologies into the fold that are still amazingly effective, but they're now being used in new ways you would've never imagined before generative AI came on the scenes to make that better. And there are very few companies that can blend all those things together like we can, just because Viya has all this stuff in a toolbox.

>> Being able to persist them and then reuse them, that's huge. We all need that.

>> It's funny, Dave, we've always talked about from theCUBE over the past decade. When I broke into the business 80s, had a computer science degree, the career was called software engineering. Engineering was in the word.software developer, then it became coder, then all kinds of other stuff started happening.

Jared Peterson

>> Just keep the personas start exploding. The aperture gets bigger. But we're coming back to software engineering, but you don't have to be a software engineer to do it. You're engineering the system. So you guys have to think about, this get your reaction on, how to enable that to be better for the user. Now the engineer, the business users, now engineering systems, they're putting stuff together.

Jared Peterson

>> That's right.

>> And so you have to be smart about knowing what they might want to need without being a coder. I think that's the trick. Can you share your vision on how that evolves?

Jared Peterson

>> Yeah.

>> I see this happening with this product.

Jared Peterson

>> No, for sure. First of all, where I'd start is I like to remind people, I talk to customers about this all the time. It's this question of, what you want to be when you grow up? And most of our customers, they don't necessarily wake up every day wanting to be a professional software development organization. They're in financial services or health and life sciences or manufacturing. And so SAS, we have to take that on. We've got to bring the professional software development, software engineering architecture and those of things to bear. And so what we've got to make sure is, as we bring tools like Workbench into the Viya ecosystem, as we continue to evolve the Viya ecosystem, that we're providing the world-class architecture for things like, how do you do generative AI at scale? How do you bring things like a prompt catalog into a workflow to make the generative AI experience better? And I take a lot of pride in the fact that we can do that because of SAS being around

>> And how do you customize it for my specific use case, for my industry, all stuff that we've been talking about for months and a year now?

>> I brought up this line of questioning because your title's platform engineering. Now, that word is used a lot in cloud-native. Rob Stretcher, our lead analyst in this area, he who's here. We were talking some of your folks about this. That platform engineering in cloud-native serverless container worldis very important.

Jared Peterson

>> Yeah, absolutely.

>> However, it's not trivial.

Jared Peterson

>> No.

>> The skills gap comes in. I want to get your thoughts on how Viya and now AI could elevate, I won't say... Let's just see level one engineer IT person, level three super platform engineering. It's hard to get that training or find people to do that, but we're seeing AI turn level one talent into level three talent across the board, from DevSecOps all the way up business analytics.

Jared Peterson

>> Yeah. No, for sure. Yeah. And we see that internally even as we're adopting generative AI tools in our software development process. Yes, absolutely. Especially in areas where whether it's in software development in DevSecOps, anything where you've got some amount of boilerplate code and those things that are part of the process. Generative AI is excellent as far as injecting into that and leveling up. And so from the platform engineering perspective, it's our job at the SAS side to make sure we're bringing the services, the tools, the capabilities inside of Viya that help people access those things.

>> Now, I got to tell you. Well, it's not a confession, it's more of an observation. When I saw the demo and the presentation, I had a throwback to 2008 where I first used AWS. When I first used AWS, I was working on a startup idea, didn't want to buy servers, that would've cost me about 10,other stuff, all that stuff. And to host it. The interface, there was no custom domain support at that URLs, there was URL. Matt Garmin laughed because when I was talking to him about the URL to right scale to do that.Anyway. But you guys have that all in one package. I can spin up resources. There's no friction.

Jared Peterson

>> No, exactly.

>> And I'm imagining that you guys must be thinking about the enablement there. Knowing your customer base, knowing what that's going to do, it's going to have that Amazon web service effect where the first time you use it, well, the alternative was provision, grunt work.

Jared Peterson

>> That's right.

>> All this heavy lifting.

Jared Peterson

>> Deploy, install, all of that. Yep.

>> So you've got this dynamic going on. What you seeing as proof points? Is it going that way, one? Because it seems like it is. you agree, but What are you seeing that's validating that? And how do you see it evolving for your customers? Because you know more about your customers than I do, I imagine it'd be huge.

Jared Peterson

>> Yeah. Frictionless is a great phrase. As we talk to our customers that are trying... They want to be doing modern data science. They want to access, whether it's generative AI or just modern algorithmic techniques, but the barriers that's in the way is whether they're even open source, whether they're SAS. like, how do I provision, how do I deploy? How do I configure, how do I tune? How do I monitor all of those things? In the cloud, you've got to have all those skills in-house if you want to do all those things, walk those steps. And so Workbench is like, Hey, let's just clear all that noise. Get out of the way.And if you're a developer, you don't need an IT shop setting this step up behind you. Just log in, click a button, it's there, start coding, run. You're good.

>> And you could just turn up and start playing with data. And then that's where the discovery comes in. Would I'd be curious to see is, and again just me riffing in real time here on this, is that because it's self-service, what accidental on purpose things happen?

Jared Peterson

>> Sure.

>> Okay, I'm going to play around, oh my God, discovery. Because we're seeing with generative AI that if you do a good prompt, you get to a quick idea or value faster, that then another prompt can follow. So you're in a time series of discovery.

Jared Peterson

>> That's The innovation cycle just keeps turning. Yeah. Well, you guys know this, but from a software development perspective, when you're building a product, when you give somebody a tool and that type of thing happens on the other side of it. There's nothing more rewarding than just seeing, oh man, okay, we gave them this thing, quick startup, generative AI. It's all those tools are in the box, and now they're just cycling and cycling and cycling. It's like, oh, I never thought that would happen.

>> Proud papa.

Jared Peterson

>> Yeah, exactly.

>> All right, bottom line, what's the bumper sticker for the product, benefits the customers, people watching? Viya Workbench, it's an on-demand, self-service compute... Self-service on-demand compute architecture or product that you can explore data quickly. I think I summed it up okay.

Jared Peterson

>> Yeah, you did a pretty good job.

>> What is the bottom line value proposition to the customer?

Jared Peterson

>> Yeah. Quick access to a cloud native, cloud scalable development environment using modern IDEs, VS Code, JupyterLab. SAS code is there if that's what you want. Pure open source is in the box, if that's what you want. If you want to blend those worlds, you can do that too. And we'll just keep adding things like R and other goodness into the box in the future.

>> And what is the benefit and impact to the customer?

Jared Peterson

>> Rapid access to actually doing real data science, delivering value in their organizations, getting from questions to answers rapidly.

>> Sure. Thanks for coming on theCUBE. Really appreciate you. Congratulations. Great product.

Jared Peterson

>> Thank you very much.

>> It's going to be big, I predict. It's pretty obvious to predict that.

Jared Peterson

>> Hope so.

>> Thanks for coming on. I'm John Furrier, with Dave Vellante. We'll be right back. Day two, we're on day two of wall to wall, two days of coverage. We'll be right back. Thanks for watching.

Jared Peterson | SAS

search

>> Welcome back, everyone, to theCUBE's live coverage here at SAS Innovate 2024. I'm John Furrier, host of theCUBE. My co-founder and co-host, Dave Vellante, extracting the signal from the noise, report on all the top stories. Go to siliconangle.com, check out all the stories. Of course, thecube.net for all the viewers, as we're in day two of two days of coverage. We got Jared Peterson, senior VP of platform engineering at SAS. He was on stage where he did the demo and the announcement of the general availability of SAS Viya Workbench. Welcome to Back to happy to be here. You'd be hard-pressed to find something I'm more excited to talk about than Viya Workbench.

>> One of thing I love about the product, and just where it's going, is it's really teed up perfectly for both the AI wave and developer needs. If you look at the market, and we report this on SiliconANGLE and cover it on theCube with other experts, the rise of open source in the AI world on the model side is merging and almost connecting with the proprietary models from size and functionality. And also the requirements to run them are getting easier, you're seeing some models running on an iPhone.

Jared Peterson

>> Sure.

>> We just talked to Intel about cost of ownership. So the developer community is booming.

Jared Peterson

>> That's right.

>> So having a Workbench is a pretty big deal.

Jared Peterson

>> Yeah, Workbench. You talk about the developer community and the rise of open source, and you can really see Workbench as us trying to reach back with our deep legacy in these areas, but then bring some of that forward and say, Okay, can we bring a tool to the market that blends both the world of SAS, SAS programming, but also if you want to just do pure open source, or you want to use open source languages like Python, but use SAS underneath the covers. Workbench just brings this all together in this amazing package. And we're really proud of it just because of all the things

>> What's the big highlight? What's the big table? People watching right now, they're SAS customers, or a prospect. First of all, analytics products well-known, and decades of success. What's the big deal? What's the high-order bit of the news?

Jared Peterson

>> Yeah. I would go through a couple things. First, I think about just Workbench is this cloud native, cloud scalable, rapidly accessible development environment. Then, once you have a development environment open, it brings the best of breed tools from the open source community to your palette, so to speak. So you've got Visual Studio Code if you want to go that angle. You've got Jupiter Lab if you want to go that angle. And so, we're bringing those IDEs to the table. Cloud scalable, cloud native, got these IDEs, but then we're bringing SAS analytics into all of that. If you're a SAS coder, the SAS language works beautifully in both of those editors. If you're a pure open source developer, Python is there underneath the covers ready to go right when you start up, you can use it in both of those editors as well. But, then if you want to say, okay, I'm interested in the SAS analytics, the reliable analytics that SAS has been known for for years, but I want that in Python. We've now got native Python libraries that are also there in Workbench. And then the last thing I would say for the SAS coder is, in Workbench we've taken all of these analytical investments we've made in the last few years in Viya that are in what we call our CAS server, via what we call actions. Things like advanced computer vision algorithms, advanced natural language processing, neural networks, those kind of things, deep learning. And now, all of that's being exposed in the form of prox. And prox are the unit of value in the SAS language. So all that capability that's been in Viya is now in the SAS language. So SAS 9 coder can come into Workbench. They know the SAS language, they get all this new goodness, and it's all there in this environment that's just lightening fast and starts up Workbench fills a nice hole in the portfolio.

>> And I'm interested in how it can be used as a lever to add incremental value to the roadmap. What can we expect now that that hole is filled? And what kind of flywheel effect can you get in the roadmap?

Jared Peterson

>> Yeah. When we look at the roadmap and you see us doing some things in Workbench that are absolutely hints as the things that are coming in the future. One of the things we like to think about in Workbench is it's like this leanest, meanest rendering of the Viya platform as possible. We've taken out anything and everything that you don't absolutely need for this pure developer persona, this pure SAS and Python coder. And so we've just leaned it out. And so you'll see us do more of that we have brought people to Workbench. They're then going to expect that experience of, man, this thing starts up fast. It's blazing fast when I'm executing code. The way we got that experience is because we leaned all those things out. So now let's say, okay, maybe they want some visualization capabilities on top of that. Maybe they want model management on top of that. Well, we've got to make sure that we also render those things out in the same way we've done Workbench so that you get all that the same. It starts up in seconds, it runs super fast. And so, yeah, it is that bridge. It's like, okay, let's bring you in. Let's get you connected to all these things, all this goodness. And then, let's bring the rest of Viya to the table.

>> And you got to keep it lightweight though.

Jared Peterson

>> That's right.

>> So architecturally, how do you do that?

Jared Peterson

>> Well, the way you do that is you only start up, you only bring to the table, and from a compute and resource perspective, the things that the user absolutely wants. The things that they're demanding. Things aren't just running all the time in the background. And so maybe they start with code, the pure Workbench experience now. And maybe they're using code for ETL, they're getting data in, they're transforming that data via SAS or Python. But then at some point, maybe they're ready to do some exploratory analysis of that data. And so at that point in the future, it's like, okay, as soon as I have that thought, I want to do this. That's when it's like, okay, let's now just go spin up the containers behind the scenes that are necessary for that experience, and only that experience, so that then get that same snappiness and performance.

>> What are some of the customer early adopters saying about the product? Obviously, I liked the UI. You guys nailed that. Great job on that. It looked easy to use. I haven't used it, but demos you guys gave, I was blown away by that. So user interface, check. I'm assuming they like it.

Jared Peterson

>> Yeah.

>> What's some of the other feedback?

Jared Peterson

>> Well, on people like it question, the first thing to note is it happens over and over again, both internally at SAS, but then externally as we interface with customers. The first time people interact with Workbench, before we even get to feature function, they just smile. It's just this delightful experience. And so that's, as a developer inside of an RD organization, that's always just fun. And you've got something when you just see this joy wash over people. But then beyond that, the thing that we hear people tell us is, you see this light bulb come on of, Wait a second, in the same platform I can enable SAS code development. I can enable pure open source development. I can enable open source languages using SAS math, and in the future other additional languages like R.When they see that they can do that, all that in one tool, that's when they're like, Oh, wow, okay, this is different.so that's the biggest thing we hear is people realizing they can bring all that underneath one offering. And so when that light bulb comes on, people get you said R is coming, yes?

Jared Peterson

>> Yes, absolutely. Later this year, yeah.

>> One of the things we see is you guys always had been talking about democratization.

Jared Peterson

>> Right.

>> Now you have a real developer product. And so I was intrigued by the slide that said, Tool for the data science developer.The data word data science developer. Now, in this world of the population needed for data scientists doesn't need to be big.

Jared Peterson

>> That's right.

>> I won't say shrinking, but you don't need to have these alpha nerd geeky data scientists jamming all the time as the table stakes. You can hire one or even use a managed service. Now you have the more user base coming on, which opens up to the developer. Classic app developer.

Jared Peterson

>> Right. Right. Yeah. No, exactly. Generative AI coming into tools like Workbench, whether it's through co-pilots and code assistants and those things is going to enable people maybe coming from more of just a developer background to then dabble in data science. So that's one way to go. And it'll be interesting to see how generative AI across the board allows developers, data scientists to level up in whatever dimension they're interested in.

>> Jared, one of the things that came up in our analyst roundtable this afternoon was, the recognition that the old days, the tech genius builds an algorithm. They company goes public. They have core technology. New technology is ubiquitous, it's now the glue that holds things together, and everything around is the new IP. And of course, we've been saying on theCUBE workflows and data new intellectual property. Which is tied into your theme. But now when you start opening up the democratization to developers to code with Workbench, the rock star, the superstar, the person could be business user.

Jared Peterson

>> Absolutely.

>> It's like, now it's shifting to not getting a little anything for the team award or a little pat on the back, or a little bump in promotion or raise. Someone in the org could be a game changer on the business model for the company.

Jared Peterson

>> Oh, for sure.a huge power dynamic.

Jared Peterson

>> Absolutely. One of things I like to remind people of internally at SAS is, whether it's with tools like Workbench or just what you're seeing happen across the space with generative AI, is that, yeah, it doesn't have to be a developer, it might be a business user. But I like to remind people, in the future we will be competing with companies we don't even know exist today because these tools enable this wave of productivity that we've just never seen and never imagined. And so, teams of one, teams of two, teams of three are going to be able to do things that we would never thought possible years ago.

>> And I the intelligence being built and devised, another one I like. Another announcement that... First the models, selling lightweight models. That's a home run, by the way. I think that's going to be excited about it.

>> The prompt saving concept is interesting, because that is a nuanced point, but it brings out the direction of saving stuff and using it later.

Jared Peterson

>> That's right.

>> And prompt-less potentially around the corner, a prompt-less environment. It just does it for me.

Jared Peterson

>> Right, right. Yeah, no. We're really excited about the prompt catalog work. And one of the reasons we're so excited about it is it allows us to blend both new generative AI, which of course is dependent on prompts, that's the way you interact with that system. But a prompt catalog allows you to bring what you might consider to be more traditional NLP task and technologies into the fold that are still amazingly effective, but they're now being used in new ways you would've never imagined before generative AI came on the scenes to make that better. And there are very few companies that can blend all those things together like we can, just because Viya has all this stuff in a toolbox.

>> Being able to persist them and then reuse them, that's huge. We all need that.

>> It's funny, Dave, we've always talked about from theCUBE over the past decade. When I broke into the business 80s, had a computer science degree, the career was called software engineering. Engineering was in the word.software developer, then it became coder, then all kinds of other stuff started happening.

Jared Peterson

>> Just keep the personas start exploding. The aperture gets bigger. But we're coming back to software engineering, but you don't have to be a software engineer to do it. You're engineering the system. So you guys have to think about, this get your reaction on, how to enable that to be better for the user. Now the engineer, the business users, now engineering systems, they're putting stuff together.

Jared Peterson

>> That's right.

>> And so you have to be smart about knowing what they might want to need without being a coder. I think that's the trick. Can you share your vision on how that evolves?

Jared Peterson

>> Yeah.

>> I see this happening with this product.

Jared Peterson

>> No, for sure. First of all, where I'd start is I like to remind people, I talk to customers about this all the time. It's this question of, what you want to be when you grow up? And most of our customers, they don't necessarily wake up every day wanting to be a professional software development organization. They're in financial services or health and life sciences or manufacturing. And so SAS, we have to take that on. We've got to bring the professional software development, software engineering architecture and those of things to bear. And so what we've got to make sure is, as we bring tools like Workbench into the Viya ecosystem, as we continue to evolve the Viya ecosystem, that we're providing the world-class architecture for things like, how do you do generative AI at scale? How do you bring things like a prompt catalog into a workflow to make the generative AI experience better? And I take a lot of pride in the fact that we can do that because of SAS being around

>> And how do you customize it for my specific use case, for my industry, all stuff that we've been talking about for months and a year now?

>> I brought up this line of questioning because your title's platform engineering. Now, that word is used a lot in cloud-native. Rob Stretcher, our lead analyst in this area, he who's here. We were talking some of your folks about this. That platform engineering in cloud-native serverless container worldis very important.

Jared Peterson

>> Yeah, absolutely.

>> However, it's not trivial.

Jared Peterson

>> No.

>> The skills gap comes in. I want to get your thoughts on how Viya and now AI could elevate, I won't say... Let's just see level one engineer IT person, level three super platform engineering. It's hard to get that training or find people to do that, but we're seeing AI turn level one talent into level three talent across the board, from DevSecOps all the way up business analytics.

Jared Peterson

>> Yeah. No, for sure. Yeah. And we see that internally even as we're adopting generative AI tools in our software development process. Yes, absolutely. Especially in areas where whether it's in software development in DevSecOps, anything where you've got some amount of boilerplate code and those things that are part of the process. Generative AI is excellent as far as injecting into that and leveling up. And so from the platform engineering perspective, it's our job at the SAS side to make sure we're bringing the services, the tools, the capabilities inside of Viya that help people access those things.

>> Now, I got to tell you. Well, it's not a confession, it's more of an observation. When I saw the demo and the presentation, I had a throwback to 2008 where I first used AWS. When I first used AWS, I was working on a startup idea, didn't want to buy servers, that would've cost me about 10,other stuff, all that stuff. And to host it. The interface, there was no custom domain support at that URLs, there was URL. Matt Garmin laughed because when I was talking to him about the URL to right scale to do that.Anyway. But you guys have that all in one package. I can spin up resources. There's no friction.

Jared Peterson

>> No, exactly.

>> And I'm imagining that you guys must be thinking about the enablement there. Knowing your customer base, knowing what that's going to do, it's going to have that Amazon web service effect where the first time you use it, well, the alternative was provision, grunt work.

Jared Peterson

>> That's right.

>> All this heavy lifting.

Jared Peterson

>> Deploy, install, all of that. Yep.

>> So you've got this dynamic going on. What you seeing as proof points? Is it going that way, one? Because it seems like it is. you agree, but What are you seeing that's validating that? And how do you see it evolving for your customers? Because you know more about your customers than I do, I imagine it'd be huge.

Jared Peterson

>> Yeah. Frictionless is a great phrase. As we talk to our customers that are trying... They want to be doing modern data science. They want to access, whether it's generative AI or just modern algorithmic techniques, but the barriers that's in the way is whether they're even open source, whether they're SAS. like, how do I provision, how do I deploy? How do I configure, how do I tune? How do I monitor all of those things? In the cloud, you've got to have all those skills in-house if you want to do all those things, walk those steps. And so Workbench is like, Hey, let's just clear all that noise. Get out of the way.And if you're a developer, you don't need an IT shop setting this step up behind you. Just log in, click a button, it's there, start coding, run. You're good.

>> And you could just turn up and start playing with data. And then that's where the discovery comes in. Would I'd be curious to see is, and again just me riffing in real time here on this, is that because it's self-service, what accidental on purpose things happen?

Jared Peterson

>> Sure.

>> Okay, I'm going to play around, oh my God, discovery. Because we're seeing with generative AI that if you do a good prompt, you get to a quick idea or value faster, that then another prompt can follow. So you're in a time series of discovery.

Jared Peterson

>> That's The innovation cycle just keeps turning. Yeah. Well, you guys know this, but from a software development perspective, when you're building a product, when you give somebody a tool and that type of thing happens on the other side of it. There's nothing more rewarding than just seeing, oh man, okay, we gave them this thing, quick startup, generative AI. It's all those tools are in the box, and now they're just cycling and cycling and cycling. It's like, oh, I never thought that would happen.

>> Proud papa.

Jared Peterson

>> Yeah, exactly.

>> All right, bottom line, what's the bumper sticker for the product, benefits the customers, people watching? Viya Workbench, it's an on-demand, self-service compute... Self-service on-demand compute architecture or product that you can explore data quickly. I think I summed it up okay.

Jared Peterson

>> Yeah, you did a pretty good job.

>> What is the bottom line value proposition to the customer?

Jared Peterson

>> Yeah. Quick access to a cloud native, cloud scalable development environment using modern IDEs, VS Code, JupyterLab. SAS code is there if that's what you want. Pure open source is in the box, if that's what you want. If you want to blend those worlds, you can do that too. And we'll just keep adding things like R and other goodness into the box in the future.

>> And what is the benefit and impact to the customer?

Jared Peterson

>> Rapid access to actually doing real data science, delivering value in their organizations, getting from questions to answers rapidly.

>> Sure. Thanks for coming on theCUBE. Really appreciate you. Congratulations. Great product.

Jared Peterson

>> Thank you very much.

>> It's going to be big, I predict. It's pretty obvious to predict that.

Jared Peterson

>> Hope so.

>> Thanks for coming on. I'm John Furrier, with Dave Vellante. We'll be right back. Day two, we're on day two of wall to wall, two days of coverage. We'll be right back. Thanks for watching.