Evolving the Enterprise

Navigating the Future of AI and Analytics With Sonny Rivera, Senior Analytics Evangelist at ThoughtSpot

• SnapLogic • Season 3 • Episode 8

Share episode

In this episode of Evolving the Enterprise, host Dayle Hall sits down with Sonny Rivera, Senior Analytics Evangelist at ThoughtSpot and Snowflake Data Superhero. They delve into the evolution of analytics and business intelligence, the impact of generative AI on the industry, and the ethical considerations surrounding AI implementation. Sonny shares his journey from software engineering in the defense industry to becoming a thought leader in cloud data analytics and offers insights into the future of AI integration in business.

Tune in to explore the future of AI and analytics with Sonny Rivera and gain valuable insights for navigating the evolving landscape of technology and business.

Sponsor
The Evolving the Enterprise podcast is brought to you by SnapLogic, the world’s first generative integration platform for faster, easier digital transformation. Whether you are automating business processes, democratizing data, or delivering digital products and services, SnapLogic enables you to simplify your technology stack and take your enterprise further. Join the Generative Integration movement at snaplogic.com.

Additional Resources

Navigating the Future of AI and Analytics

Welcome to Evolving the Enterprise, a podcast that brings together thought leaders from the worlds of data automation, AI integration and more. Join SnapLogic’s chief marketing officer, Dayle Hall, as we dive into the captivating stories of enterprise technology successes and failures through lively discussions with industry-leading executives and experts. Together, we'll explore real-world challenges and opportunities that companies face as they reshape the future of work.

Dayle Hall:
Hi, and welcome to our latest podcast of Evolving the Enterprise. I'm your host, Dayle Hall, CMO at SnapLogic. Today, we're delighted to welcome a true luminary, and that's a big word and I don't use it lightly, but he is a luminary in the world of data analytics, Sonny Rivera. He's the senior analytics evangelist at ThoughtSpot, a company that many of us know pretty well. He's been instrumental in pioneering advancements in the data landscape. He's recognized as a Snowflake data superhero. So I'm sure we'll get into that shortly. Vast experiences with big companies like Ally Financial, Hartford, AAA, have solidified his position as a thought leader in a cloud data analytics platform. So super excited to have him on today. Beyond just the corporate accolades, he has a very entrepreneurial spirit, we'll probably get to some of that today, and he's the founder of two companies. And for those eager to dive further into the world that Sonny frequents and has his expertise, Sonny, it's an absolute honor and a pleasure to have you on today. Thanks for joining.

Sonny Rivera:
Well, it's wonderful to be here. And gosh, just to hear all of those things, it's not all that great, but thank you very much, those are very kind words.

Dayle Hall: 
No, I've looked, you're very accomplished. I think what I usually like to start with is just give us a couple of minutes, how did you get into this world, what brought you ultimately to ThoughtSpot, what are you trying to achieve there, but give us some of the background and how you actually got into all this experience.

Sonny Rivera:
So I'm going to go way back, way, way, way back. If you're old enough, you could remember these TV-guided weapons that we saw in the first Gulf War, that was steering these weapons, that was me. I wrote the software for the guidance system, for those systems, right? And so I was always involved with software engineering first. I mean, I was a software engineer and processing data, doing image processing. And being in the defense industry, I learned a little bit about how important process was and helped them get to that mature process at what would be at the time, SEI Level 5, right? Yeah, the one, two, three, four, five levels in the capability maturity model, so I learned a lot about that. I helped them get there and define those processes. That made me think, you know what, this industry really does need to mature. And so we went out there and we started our own company around business process modeling, data modeling, business process modeling and code generation. 

And so I did a company there, got out of that around 2000. And what happened at that point was I said, I really ought to just go back into industry, and I did and I got to work for some great companies and building those kind of data platforms. And I loved it so much, I said, I really need to tell other people about this. And I just started telling people. I was running my own user groups and doing all these things. And I see, I'm not getting paid for this, so I get up. I got some swag and ThoughtSpot came along and said, you know, you're doing a really great job at this. We love your expertise and we'd love for you to share that with the rest of the world. And so here I am. 

Dayle Hall:
That’s good. Well, so you got definitely some deep and yet varied experience. Given some of your early experience around the missile system, that's really interesting. So we're going to break it into two or three different sections. We’re going to, first of all, start with absolutely in your wheelhouse, but talk about analytics and BI, how it's evolved and where we're going in the future. And my first question is, and something that I think you said is this concept of the fifth wave. Maybe you can explain that to us, how does this fifth wave of analytics and BI innovation differ from what we've seen before and what are the things that you're really excited about, what are the things that you see as opportunities for businesses today? 

Sonny Rivera:
Yeah, so the great thing about this, we put out- ThoughtSpot, me and some of my other cohorts there, Cindi Howson, Benn Stancil, we put out trends and predictions every year and we put out these tenants. So we collaborated together on these. And one of the things we've seen is early on, that first wave was really about ad hoc analysis, enabling developers to write reports without having to write SQL, and that took probably 15 years to mature, right? And then we saw visual discovery tools, and that took another 10 years to mature, tools like Qlik and TIBCO and Power BI. 

ThoughtSpot came along in the age of- and defined an age of augmented analytics where it's natural language processing focused on business users, really making it easy for users. And then we saw the modern data stack, right? And we saw companies like Looker and others come along. And then even as the modern data stack hasn't matured, we've seen this generative AI, this fifth phase that's come out, right? And there's so much excitement around this particular area. And I think what we're going to see in this fifth wave is more multimodal type of analytics. I think you'll see analytics that goes- how should I say it? Really, what I think you'll see is analytics become less visible or invisible even as insights become more impactful. You're going to see embedded analytics into back-end business processes, customer-facing portals, partner portals, all of that embedded analytics, that it's super easy, whether it's on your phone or on your watch, or you can talk to it or write to it. 

Dayle Hall:
Yeah. You mentioned something that I thought was interesting, you mentioned the requirements, the advent of business users. Do you think that- look, technology develops at an exponential rate these days and sometimes technology is developed and it's kind of looking for a home, but do you feel like some of the requirements from the business users, the lines of business, what the data we need to make smart decisions in a line of business, has that driven this latest wave, or was it just a natural and accelerating progression of the technology? 

Sonny Rivera:
Yeah, I think it has, but I do think there's been a trend in that way, going all the way back to 2000, where we had a lot of tech driving things, the technology teams, CTOs were driving things. And I think the businesses have made their mark and said, we want control back, we want to make sure that we're doing things that drive business value. And that's just been amping up and amping up and it's gotten even stronger with generative AI. How can we use this, how can it make us money, how can it save us money, how can it derisk our platforms and our operations? 

Dayle Hall:  
Yeah. And one of the things- I'm going to get into a couple of generative AI questions in a second, but one of the things that we hear from customers, I'm sure you do too, is this question now, is your infrastructure, is your data- I don't necessarily like the term AI-ready, generative-AI-ready, but what do you think- what's stopping organizations today from being ready to take advantage of AI, generative AI specifically, and how can businesses prepare to make sure that they can take advantage, are there some fundamentals they have to have in place?

Sonny Rivera:
Right. Well, you're in this data space, data pipelines and you know data quality has been an issue for as long as we've had data warehouses and we've had data analytics. And it continues to be a real problem in the industry. So I think having quality data is what is holding companies back. Even if you haven't selected a technology, a partner, a platform to work with, or what you're going to be doing, you need to be improving the quality of your data so that you can integrate it in. And then the second part that's happening, generative AI like, hey, I use this LLM from Google or from OpenAI can hallucinate, can give you may be biased information, but if you can use a newer approach, like RAG, which is a retrieval-augmented approach, where we'll retrieve data from your dataset and let the LLM do what it does best, which is generate text, but use the information from your data, now you're in business.

Dayle Hall:
Our CEO just put a blog out last week, which is about RAGs to riches- 

Sonny Rivera:
Nice.

Dayle Hall:
And that joke I’ve heard in a while, but yeah, I understand what you’re saying there. Do you think that- I mean, as you're talking to organizations, do you feel like they're fully embracing this opportunity? Are they looking for ways to improve the quality of the data? Our businesses, sometimes you can be hamstrung because they don't actually know where to start? Do you see that as you're talking to ThoughtSpot customers and people in general in the industry? Is there a, we have to move but we don't know where to start, or are people actually embracing it?

Sonny Rivera:
I think people are embracing it. I was reading a study recently. I think it came from E&Y, from a survey they did, 99% of the CEOs said they are moving on and strategizing for LLMs, for generative AI already. So of that survey, they are doing it, but that doesn't necessarily mean they know what they are doing and how to do it properly. So I do think it is important. If I were to advise them what I’d give to customers and clients, data quality, take a look at that. Focus on business problems that you can actually implement real actions on. And so when you think about its heart- let me just tell you and step back. At its heart, analytics is about making good quality data, identifying insights and then acting on those insights to realize value. If you do those three things, it doesn't matter whether it's GenAI, if I'm doing it in the back-end, if I'm automating it with a pipeline type of tool, you’ve got to do those three fundamental things and that's where I would start. 

Dayle Hall:  
Yeah. And I think, as you said, no matter if you're using other other elements of AI, or just in general, it doesn't even matter if you're using generative, but the principle of having a good analytics organization or tool or output is actually to make better business decisions. Do you feel like with generative AI- and we talked about business users, do you feel like the business users have the right questions, or they know what to ask regardless of the underlying infrastructure that you have for the data? It's good that you said 90% of CIOs from that survey are embracing it. Are they getting overwhelmed yet? Because everyone is talking about it, everyone is asking what they can do. Are they ready for this?

Sonny Rivera: 
Yeah. Again, I'll go back to I don't think that they're ready for this at the moment. And I will say you've probably been reading my blog because I just wrote a blog about it. It was entitled, Are You Asking the Right Questions? And I think the big point is people- and I wrote that because when I talk to customers, they're often asking me, hey, what do you think about this technology? And I respond with, well, what problem are you trying to solve? Is that even the right question? And so we get to, hey, do I- am I measuring the right thing? Things as simple as how much transparency do you need, can this be completely a black box, or do you need to know from a compliance perspective all the way through, do you need to understand your data quality, are you going to use your own data in this? 

So I think there's a series of questions that people need to be asking and they'll vary by company and by organization. But making sure that you're asking the right question before you leap in is critical to any infrastructure change, any product that you're going to- any initiative you're going to take forward, do those things first. 

Dayle Hall:  
Yeah. And it's been crazy to see how fast generative AI has taken off. On this podcast a few seasons ago, one of the things that a few of the guests talked about was, let's say you're trying to sell us a software that's AI-driven, again, before generative, their guidance on the podcast were you still have to be transparent about how it works, what it does, where the data is coming from and so on. You just used the term that I was thinking of, which is the black box. You can't just have a black box when no one really knows how it's being used, where it's coming from and so on. And that's going to lead into my next question, which is around ethics, but has generative AI muddied the water around ethics and using AI, or has it made it clearer, in your opinion?

Sonny Rivera: 
I definitely think it's muddied the water a good bit. Everything from can it write my term paper, can it write my college essay, to can it write my email to a customer, the response to a customer or write a tweet? So I think you see all of those things happening. I do think it's important to have transparency. I think it's important to have trust. And trust is an interesting concept, I'll come back to that. Having the ability to have traceability, transparency and a human in the loop is critical. There's a couple of terms we know here, right? One is human in the loop. And I'll say that means, hey, before anything happens, a human is actually okay-ing it and moving it forward. 

There's also the concept of human-on-the-loop, right? And that means they basically get the big red button to press and say stop. So a good example of this would be imagining that you had robots that were, I don't know, cleaning your yard and you just let them do it. Well, all you are is you're on the loop. Something happens, it looks bad, they're running out into traffic, they're banging on cars, what do you want to do? You want to press the big red button and stop. But if you're human-in-the-loop, the robot says, hey, do you want me to cut that patch of grass in the front, do you want me to do that bit of work? So that's kind of in the loop, or is this right? Before I go forward, this is what I'm going to do. Is this right? Yeah. I think you need that human-in-the-loop in the analytics world before we're ready to go full out, I'm just human-on-the-loop, we'll let everything run forever. 

Dayle Hall:  
Yeah. Obviously, I've heard human in the loop. I like the human-on-the-loop concept. Because you just mentioned trust, I'm going to ask you, if the processes- if you have that human involved, does that deliver the trust,  or are there other things that we need to be thinking about?

Sonny Rivera:
No, I think one high-quality data delivers that trust. I do think we definitely need to make sure that we're including more people. And again, we've had this problem for a long time, breaking down silos, it happens to build that trust. I also think that with the introduction of more advanced AI, GenAI, we have to understand what does it mean between you, a stakeholder, me, an analyst, and I'm using AI. We've had a trust problem as analysts anyway. One piece of data goes wrong, what do you do? Oh, I can't use any of your numbers. Now, let me add in the fact that I'm using an AI in there. So I think rebuilding that whole trust model of, what does it mean for us to work together and include an AI with it. And again, you get back to traceability, transparency and that human-in-the-loop.

Dayle Hall:
Yeah. Again, on some of the people we've had on the podcast, they represent organizations that are very closely linked, or are actually driving the ethics discussion and responsible use of AI. And what I said to them at the time was, I- because I have kids. I have a 16-year-old and a 13-year-old. And what I'm seeing now, they're going to be like the things that they're going to see when I'm gone it's just going to be crazy. But I feel like with people that are really trying to stay on top of the ethics piece, I feel a little bit more comfortable that we're kind of attacking it the right way. And I use it a little bit as the advent of things like social media. 

I know it's very different, but I felt like the social media sites just hit so fast and no one really thought about a lot of the negative implications. So we were kind of then trying to put the horse back in the stable. I feel like we’re thinking about the ethics piece a little bit more on AI. Where do you feel like people like yourself, who have big impact and a lot of knowledge, and other brands that are potentially driving some of this technology, where does the responsibility lie? How should we get involved with the ethics piece and how do we make it safe enough that we don't have to worry about some of the scare tactics?

Sonny Rivera:  
Well, I think, one, the ethical use of AI is really everyone's responsibility, all the way down from the consumer to the producer of the service as well as those implementers behind the scenes. So think about if you're going to license a technology, a platform, then part of your evaluation criteria has to include how are they implementing ethical AI and giving you opportunities to control that and manage it according to how you want to run your operations. The same thing internally, I think you've got to educate your people as an organization. So back to data literacy, AI literacy, what does it mean to use these and what are those ethical questions that we face, whether it's bias or [threatening] or do we have the actual legal rights to use these things. So I think understanding and educating the groups there. And then lastly, as consumers, we need to be able to say, what are we ethically doing with this information and is this actually trustworthy in and of itself? 

Dayle Hall:  
Yeah. I think I like that last point specifically around it's also incumbent on us to be part of the discussion. Because again, and I know it's very different to social media, but I felt like we all jumped on that and then it was only afterwards, we started to think about, oh, maybe some of this is not the best and we need a different way to control it or safeguards and so on. So look, it's good feedback, good information. I want to move on to an area that I know you're passionate about  and we talked a little bit about in the pre-meeting before we had this podcast, about financial operations and things like LLMOps, that's a mouthful, some marketing person created that acronym that's for sure. Anyway, could you just explain to some of the listeners on this, what is the transition from financial operations to LLMOps and why is that significant for an organization when they're managing AI and machine learning?

Sonny Rivera:  
So a couple things. One, with the modern data stack and consumption-based models, companies, a lot of them got burned because they didn't fully understand the models between OpEx and CapEx. They didn't understand the consumption models and maybe they got burned, but they now know, they've learned, right? They know the good, the bad and the ugly of these consumption models and they're also putting in the best practices. And so something like 81% of enterprises now have a FinOps group within the organization, maybe not an LLM FinOps group, but just a FinOps group. And so they're extending what they've done with some of their other services to that LLM space. And I do think it rolls up all together. And you're going to see companies that are facing these cloud services bills where they're getting unexpected charges for things obviously unexpected, right? Or they're getting surprise bills where, frankly, I've done that where, hey, we use this service, not a lot of governance, spun up the services and then we went to sleep for the weekend or the next two weeks and we saw these massive bills. We've all heard those stories, but they've been burned by that and they're going to put these guardrails in place now. 

Dayle Hall:
Do you think- I don't want to say, I don't want to be completely declarative, but do you think that consumption-based model, not that it's dead, but are people more aware of it now? Are there specific concerns around managing LLM capabilities internally that people are aware of? Or are we going to go down the same path of, okay, we had consumption, we've made some mistakes, but we think we're going to be fine with LLM? Are we just on the same path?

Sonny Rivera:  
I think we've skipped ahead. I think we know the best practices. So I don't think consumption-based models are dead at all. If we manage them, well, they're very, very efficient. So let's make sure that we're managing them well. And I think enterprises know that today. And that's why they're trying to fold these in. I would say what's really, really new here is the pricing models for these services is that it varies greatly across the different platforms. So it's, again, hard to compare so you have to do your homework, make sure you know those pricing models and make sure that you're comparing across them. 

Dayle Hall:
Yeah. Sage advice, Sonny. Someone is going to listen to this podcast and go, you know what I need to do tomorrow when I get back to the office? Yeah, with that, I'm sure that that's happened to more than- you are right. Away from that type of consumption model and pricing and so on, let's just talk about controlling of costs or being cautious of cost. When it comes to general innovation, particularly around, obviously, AI technologies, we think everyone, the data says everyone is getting involved, people are staying, even though sometimes they don't really know what they're trying to achieve. Is there a danger that as people start to bring in these new technologies and try and do all these support the lines of business? Are we creating another potential cost disaster because everyone now wants to do something around this type of generative AI technology?

Sonny Rivera:    
If I'm doing a talk, there's 100 people in the room, I'll ask, how many of you guys are using GenAI or LLMs? Almost every hand goes up. How many of you are using them in production? Just a couple, one or two. So do that, ask that question. I think people are dipping their toes in and they're asking those questions like, how much is it going to cost me, can I actually scale this and more importantly, is there a specific business case where I'm getting the outcome that I'm looking for? 

Dayle Hall:  
Business case, those are the words that we're looking for, Sonny.

Sonny Rivera:  
Yeah, I think it's got to drive from there. And here's what I talk to a lot of CDOs, I talk to a lot of business leaders about is, hey, let's look for quick wins but that align with your data strategy. One, quick wins are out there. They may have all been taken up because the predecessors have looked for quick wins in implementing, but you’ve got to find the ones that align with your overall business strategy and your all data strategy, that will scale. So they're out there. But doing a series of ad hoc quick wins doesn't do anything for you and your data strategy, it doesn't do anything for you and your business strategy. Try to align those things

Dayle Hall: 
Do you feel like- for the people you talk to or advise or the companies that I know ThoughtSpot are  talking to, people, do they understand that, that the business case is probably just as important and understanding what success looks like is just as important as actually getting the right data and having the right tools to take advantage of it?

Sonny Rivera:  
I don't know that we always have the right people in the room. 

Dayle Hall:
What do you mean by that? Who should be in the room?

Sonny Rivera:    
For instance, sometimes you will get, I don't know, a head of IT or a head of data that's looking at this and they're looking at it purely from the perspective of, how can you make my life and my team's life easier? Hey, could SnapLogic reduce the amount of developer time I have that we're doing and build these automated pipelines? And yes, that's all great. That is great, right? But what have you done for the business? Because that's where real scale is, if I saved you 10% on your development costs, that's one thing, it's a good number. If I saved you 10 10% or increased your sales by 10%, that's a whole different ball game.

Dayle Hall: 
Certainly, a bigger number.

Sonny Rivera:  
I will say I think having the right people in the room, this is one thing I love about ThoughtSpot, we're always out there talking to you like, hey, yes, we can help you. That's great. Do you have the right business case? Is it going to help you? Because we want to be your partner in the long run, not for that one-off thing.

Dayle Hall:
Yeah. You probably in these situations, we have, I have. I've heard about this for a long time around things like the buying group is getting bigger, you have more people that you need to kind of bring onboard and so on. But it sounds like it's almost like success is dependent on making sure that a broader group of people really understand the technology that you're looking at, the cost, some of those quick wins, how it's going to help long term on the business. It feels like whilst if you're in sales, you’re like, if I get one person in the room and that person can sign the check, I'm selling to that person. But long-term success for the business, you still need a broader group. So if someone's out there, maybe they're about to get into a sales process, or they're looking at a tool, what advice would you give them around who to bring in the room in their own company as they start looking at this?

Sonny Rivera:  
Yeah. It definitely varies, but I would say you want one to two business stakeholders in the room. Having those folks, if you're in a highly regulated area, like financial services, maybe health care, then you want to make sure that you have those architects, there’s somebody from an architecture or from a compliance perspective that makes sure that you can actually deliver what they need. There are folks in organizations that won't necessarily realize that this product doesn't meet the compliance goals, or how it will meet your compliance goals going forward. So I think it's important to look at it. 

It'll vary. Have the business stakeholders have that person that owns the delivery of that particular product, whether it's a product owner or a director have them in there, yeah, those are probably the key folks that I'm looking for. And then doing a proof of value with end users that can actually feel it and touch it is priceless, right? You'll get the buy in from there. If your product is going to resonate with them, that's where you're going to get it. And I guess my point here, I would tell every one of these folks, everything that happens magical in your company and magical in my company happens on the frontline, happens in the interactions with our customers, whether it's with our product or with our teams, that's where all the magic happens. 

Dayle Hall:
Yeah. The things I've heard from other people on the podcast and talking to customers and so on, the mistakes in putting technology or it could be any technology, it could be generative AI, it could be an analytics tool, is if you don't start with the business use case, the things you're trying to solve for the business and just say, this looks like a cool technology, it may still be successful, but you don't get the lift, you don't get the real ROI that you could. And there's so much technology out there, it's almost like just taking a little bit of extra time as you're looking for the technology and making sure you really understand the business use case, I think the technology investments will be 10 times more successful.

Sonny Rivera:  
Yeah. And when I look at this, I'm not looking at it from how is ThoughtSpot going to be successful. If I can make you successful, the rest of that stuff is going to take care of itself. So I'm focused on making you successful. If we can do that, the rest of that will take care of itself.

Dayle Hall:
Yeah. Okay, so I want to ask a question because I think in the pre-discussion for the podcast, you said something like AI might make some products worse. Can you elaborate on what you mean by that? And some of the challenges around things like user experience, maybe product quality,  what do we have to be aware of? What are we not seeing, Sonny? What are you not telling us?

Sonny Rivera:  
Using AI, building a chatbot is relatively simple. It's pretty easy to do. Throwing that into your product is pretty easy to do. But actually, building a good chatbot, a quality product is hard to do. And I do think that we are going to see a little bit of things getting worse before they get better for some products. We're going to- products are rushing to market as quickly as they can to maybe throw a chatbot on top of whatever it is they do to say I'm in the AI space. They may not even be in the AI space, they're just wrapping a chatbot in that space. And someone might say, ThoughtSpot has a conversational aspect to it. Haven't you guys just done that? And I would push back and say, one, we were a first-mover. Two, for the last 10 years, we've been in this search and AI space and we've got 25 patents in that space. What made us successful with our Sage product was how we took, we used natural language to do what it does best and transfer or translate your natural language into our proprietary, patented semantic layer to give you the answers that you need. A little bit different there, but that’s- I don't think companies are doing that same thing, it's how quickly can I put a chatbot in my product?

Dayle Hall:
I've definitely had those types of situations. So when you hear that- and again, I think your example or thoughts about doing it the right way, making sure the patents are there, making sure the go-to-market stuff will all come afterwards, but if someone else is looking at potentially integrating some of the AI capabilities into their own products, based on your experience or what you've seen be successful, how do they mitigate the risks? What are the things that they need to be really careful about not just in product development, but then potentially the sales, the marketing and beyond? 

Sonny Rivera:  
Yeah. So, one, I think you can think beyond the chatbot, right? So first of all, take the chatbot ad off. And there's more to AI than just chatbots. So I would advise folks to do that. What about metadata repositories? Think about those. You can't just wrap this chatbot around your product. You still have metadata and real data to deal with. Do you have the infrastructure in place to deal with this? What about the accuracy of the results that you're getting, the provability, the trustworthiness that you're giving? Imagine that you got a poor quality product out there, and we saw some come out early, and they said the worst and most awful things? You don't want that as your brand? And then I would say think about feedback loops to help your models, to help your product get better and learn from what's happening. 

Dayle Hall:
Yeah. As we come to the end of the podcast, there’s a couple of questions I have around the future. So, one, would just be what do you look at as potentially using AI in product development? We just talked about how people are trying to integrate AI into their own products. What about using AI to deliver better products and services? Again, look, I'm in marketing, so I can be honest, I have used some of the tools to help me write a better blog. Again, I'm still the human in the loop. So I definitely had some editing afterwards, but the things that people need to be careful about using AI to deliver better products and services themselves?

Sonny Rivera:  
Yeah. I would say one of the things that didn't make it in my prediction here was that I think you're also going to see a series of cottage companies coming up to fix AI-generated code, right? You're going to see that. 

Dayle Hall:
Right, yeah.

Sonny Rivera:
I would go on record and say that. But if you are going to do that, if you're going to use AI to help you generate code, I want you to take a look at what mode. A company that ThoughtSpot acquired is doing- we have an AI- it's more geared toward data teams that are actually doing software. So you're describing what you want to do, it's generating the code, it's got the human-in-the-loop and then letting you integrate that into your workflow. It is not talking to a chatbot to build me a data pipeline that pushes everything out to email campaigns, out to Salesforce or a marketing campaign tool. So I do think having that approach, human-in-the-loop, think about how can you make the process that the developer is going through easier, transparent and not hallucinate, that's got to be that's got to be a big part of it.

Dayle Hall:  
Yeah, that part specifically still feels like we haven't really fully understood the level of hallucinations. And I think obviously, if we all stay involved in the process, we can mitigate it. But I think there's going to be some more public examples of what not to do. 

Sonny Rivera:  
Yeah. And I'll tell you real quick,  things that I do with these chatbots is, I often ask it to generate me data for my sample test, generate me a bunch of data to do this. And I asked it things like, hey, how do you write that, whatever, that Excel function? How do you write that Python function? And so being able to incorporate those is invaluable because we've all used referenced material over and over again. Just make that reference a little bit easier. And then the last thing I will leave you with, you're probably like me, you drive a modern car. My car does a lot of things, right? One thing it does not do is it does not drive. It can assist me. It tells me if I'm changing their lane wrong. It might parallel park, I'm not great at that. But it assists me. It does not drive for me. And I think if you take that model, you'll probably be in good shape. 

Dayle Hall:
That’s a good lead into my last question for the podcast and I asked a lot of other industry leaders like yourself this question. If there was something that you look out- it could be 12 months, could be 3 or 4 years, is there something that- with some of this new capability around generative AI, is there something that you look at and you think, I'm really excited to see where it goes in this industry or in this part of my life or where businesses are going to take it? Is there something that you would leave the listeners with that says, this is what I'm really excited to see?

Sonny Rivera:  
I think when I think about that, I'm excited to see how this becomes more integrated. Like I said, more invisible earlier, I kind of mentioned that. So this multimodal environment that we will see where it just becomes seamless and we don't even recognize that we're doing analytics. We're doing what you and I normally do, we sit here and we're having a conversation and we're thinking about ideas and probing that and going down this particular, okay, now let's come back. But just making it multimodal, whether it's your phone or your voice or your text, I'm really excited to see how that happens.

Dayle Hall:
It's great. I appreciate your time, Sonny. It was a great discussion. I know we could talk for another couple of hours. For all our listeners, again, I would highly recommend you check out some of Sonny's blogs. There's some videos online as part of his Snowflake data hero recognition, I think that's the right word. Sonny, thanks for joining us today.

Sonny Rivera:  
Yeah. Thank you so much. I’m really honored to be here. I'd like to leave your guests with something, come to thoughtspot.com, that's where you can find me. But hey, come look at those trends and predictions. I can leave you with that link. But our trends and predictions are out there. We would love for you guys to read them and let us know what you think. You can find me on LinkedIn, you can find me on Twitter at rqrivera@twitter.

Dayle Hall:
Twitter or X as it’s now called or Threads or whatever the latest- 

Sonny Rivera:
Or formally known- yes. 

Dayle Hall:
Formerly known as. Well, thank you, Sonny. I appreciate your time. For everyone else out there, thanks for joining us today, and we'll see you on the next episode of Evolving the Enterprise. 

Sonny Rivera:
Thanks for having me.

As we wrap up this episode of Evolving the Enterprise, we want to extend our gratitude for joining us on this exploration of enterprise technology. Keep the conversation going by subscribing, rating and sharing our podcast. Together, we'll shape the future of work. Until our next episode, stay innovative and stay tuned.