In the Hot Seat with Jay Gopalakrishnan

Speaker A: Insurance Unplugged in the Hot Seat where the complex world of insurance is laid bare. Hosted by Lisa Wardball, this podcast promises an unfiltered glimpse into the industry like never before. Each episode invites you to listen in on the candid conversations that usually happen behind closed boardroom doors. From deep dives with industry leaders and thought leaders to innovative discussions with minds shaping the future of insurance, we bring the most genuine talks directly to your ears. Our guests take the hot seat alongside me to explore the inner workings, challenges and triumphs of the insurance world. If you've ever wondered what goes on in the shadows of the insurance industry, from the boardroom banter to the behind the scenes strategies, this is your chance for a front row seat. Prepare for unguarded, enlightening and engaging discussions that cover every angle of insurance presented in a way that's both insightful and accessible.

Speaker B: Welcome to the conversation.

Speaker A: Welcome to Insurance Unplugged in the Hot Seat with Lisa Wardbaugh. Welcome to today's episode of Insurance Unplugged, proudly sponsored by Iris InsurTech, your gateway to the future of insurance distribution. At Iris, we harness the power of generative AI to revolutionize data processing and decision making across the distribution spectrum. Our platform integrates GENAI to provide not just insights, but actionable intelligence capabilities, configurable workflows and dynamic form generation, all underpinned by continuous data quality management. Discover how IRIS is pioneering smarter, more efficient operations in the insurance industry, paving the way for a new era of distribution excellence. Let's dive into how GENAI is transforming the landscape of insurance distribution today on Insurance Unplugged.

Speaker B: Welcome to another episode of Insurance Unplugged in the Hot seat. This time I have with me. I'm so excited about this. I mean I'm excited about all my guests but I get, there's a few and far between that I get to get really excited about. But Jay Gopala Christian is joining me, CEO, Founder of noe and today we're going to talk AI and distribution. Breaking the rules. So Jay, welcome to the show. If you don't mind introducing yourself to our guests because I don't know if they know of you or not, but I'm lucky enough to know you and I'd love it if you would. Yeah, give a little bit of your background and just explain to them a little bit of your journey on becoming founder and and CEO nowe.

Speaker C: Thank you Liza. It's a pleasure to be on here. Appreciate the invite. So my name is Jay Gopalakrishnan, founder CEO of Noe. Noe is uh, end to end AI enabled analytics platform. It's the best way to think about it in terms of the journey. I used to run product and engineering from back in the day, many different companies including Demand Force and a few others along the way. Uh, and there was a common thread across all of them around data, centralizing data without centralizing it and so on. So that's the genesis of Noe. And basically our mission is to make companies data driven and get data driven products out to market very quickly that are AI enabled.

Speaker B: Well, I love that and most people that have been following along on my hot seat. So thank you for joining us in the hot seat day. They know that we're going to keep it real and they also know that we're building this foundation of data because all of the analytics in the world of course have to be on the proper data foundation. And Jay, as soon as I got to know you, I was not only fascinated, but I'm really, really doubling down on the criticality of decentralized data and how that decentralized data really is going to be kind of the source of fueling the future with AI, with real time knowledge, with real time analytics and with real time intelligent insights. And for a long time back in my days at reinsurers and carriers. Right. I won't say like static analytics just like aren't enough. Like we have to be able to make those insights not only actionable, but they can't become actionable if they're not able to be streamed, if the data is not able to be realized. And I, I just love that about when I started interacting with you and talking to you, it was like this light bulb moment. I'm like, uh, and I, and I. So I love the fact that we're going to have this discussion today. So let's take our listeners through. We're going to talk about AI decentralized architectures, kind of how that's breaking barriers, why real time insights without data persistence is critical. And for some of our listeners, Jay, I had to learn this, explain what persisting our data means and why not persisting data is kind of next level. And then of course we'll do a wrap up. So let's set the stage. We hear AI everywhere now. So I almost think most people like okay, AI analytics, blah. Why don't you actually break that down? Like what makes AI driven analytics unique? You've really been inspired to deliver a platform to us all that prioritizes real time analytics without persisting data. Why was that Unique. And how did you kind of come about this? Let's talk about it for our guests and maybe give them a little bit about what that means.

Speaker C: Uh, sure. I think it may have to go back to the, the genesis of Noe and before then. So when I used to run product at a, at a company that had thousands of small business, small businesses on its platform, we wanted to put something to our customers to generate, uh, analytics for them on how their stores are doing, et cetera. And a typical problem is that you have data and you know, back in, back in that company we had data in 19 different silos, right. And growing. And some of that data is semi structured, some of that is structured, some of it is unstructured. So uh, what the industry tells you to do is you take all of that data, especially back then, centralize it into a common repository and then you have transportation tools to move the data, you have a transformer, transformer tools to transform the data further. You got your data warehousing initiatives, then you got your BI tools and then putting something out there for our customers. So the whole flow was, excuse the term but so asked backwards that there's gotta be a better way to do this from the ground up, right?

Speaker B: So that's, you always love solving ass backwards problems.

Speaker C: You know, then there's this whole last, you know, a number of years there's this whole thing about modern data stack and so on and that's basically companies that get fed a bunch of money to move the data and you have companies that do transformative and data warehousing, bi and so on. So uh, it just becomes a square peg in a round hole kind of use case. And it just takes forever to get to the data and the business outcomes don't really change. You're still waiting for the data. So the, so when we founded noe, the foundation of anything we did was basically on the data side. So how do you interact with data from multiple different sources without having to put it all into a central repo? If it's in a central repo, fine, great. But a lot of times it's going to be semi structured data Mongo elastic. You have your documents, you have your relational data. How do you get a view of that data? So it's like solving that problem and then putting all the pieces together on top. And the AI bit comes into play because how you interact with the data changes with AI, but you still need that foundational layer where you have a single pane of glass across all your data across displayed sources.

Speaker B: I don't know if I ever told you this, but, like, back in my. Back in my days where I ran data teams, I was just used to, uh. So I came up. Jay. I came up from the accounting consulting side, M and A. And then I. Then I became like, head of operations. And in my operational role, I had a team of like, 50 data scientists dotted lines into me. But then I had, like, actually literally 200 data people. Data people that did ETL. And when I took over as COO, my first question was. And, um, you know, you could say it was a stupid question. I was like, why do we have to etl?

Speaker C: Yeah, I mean, it's. It's a. It's a very valid question. Even today.

Speaker B: You know, they used to tell me, well, that's just because you don't know data. And I'm like, no, I just. I think I know business, and this isn't a good business proposition. And so I started my journey, and again, it wasn't from, like, the perspective that you started it from, but I started my journey on this. As this is nonsensical. It's every time you move data and every time you have to do relational. In Jay, we did so much relational mapping. I mean, can you imagine? I had 200 people doing relational columnar mapping, and I was just like, I'm never going to be able to optimize not only the data, but my process if I keep, uh. Like, I was definitely like, the shiny ball wouldn't solve the problem. The problem was inherent in the way we were mapping, moving and transforming data. So, so what are the. And then, of course, like, if you put a reporting tool on top of it, that's great, but the reporting tool is limited by definition to the. To the. The inherent data processing and setup that you have underneath that. So how does a solution, like what you're doing at Noe and the way you're thinking about it, how does that change the paradigm of some of these traditional reporting hurdles? And what does that look like? Like, like, I think I know because, you know, but for somebody that's not technical, what does a traditional hurdle feel like as a user of reporting?

Speaker C: Yeah. So, you know, traditionally what happens is like, hey, I need some. I need some data, or I need some reports for this. Right. And that gets sent into it, and someone goes in and writes some queries, gets the report out, and do they maybe add it to a data warehouse and then run the report and then send that over? Right. That's the historical kind of view into that process. If you Think about it differently. And this is what we've attempted to do with the platform is you have a data set as a service layer and this data set as a service is basically saying, hey, I might have data in a number of sources. This data set is going to give you a view of that data across those sources that's trusted, validated. All the underlying complexity of that data and which system it is, how you're going to query it, is it real time, is there like a lag behind it, et cetera is all encapsulated in this data set as a service layer. So then it negates the need to do the centralization process essentially from there because we know the metadata and everything associated to that data set. Now it's AI enabled. So when you put an AI engine on top for us, we have to use our own AI engine on top because we can't have anything go outside of the Novi infrastructure. Then a couple of things happen. When you ask a question, it federates that question into all your data sets that the user with their rights have. Ah. And then does the calculations and give you the answer back. So it's almost like there's a mishmash of your unstructured, non deterministic world in the AI and your rules based world and then combining them together to say, hey, this is what your answer looks like and this is how we validated

Speaker A: it, which I really love that.

Speaker B: Well, no, I'm like, like at the edge of my seat because I was literally just talking to somebody on a call like immediately before our session and we were talking about like, we were talking about productivity tools and I won't name names, but I was like, yeah, but like human behavior doesn't like to look right and have like this other way that they do this. I think we were talking about email productivity tools and they were like, that's the salvation to all things AI. And I'm like, no, it's not. Because email isn't the primary. I mean, okay, yes, back in my consultant days, email was all I did, but my primary work is usually done in some sort of an, I'll call it a portal. And so, uh, what I was telling this person is productivity tools are like, honestly, okay, basic, but the least impactful thing that I'm going to do to my ROI of my users and my business stakeholders. And the reason why, Jay, that I think that's interesting what you just said is because if we think about analytics and reporting, people don't think most users don't think like well, I've got this, you know, unstructured, um, data over here and I've got this structured data over here and I'm going to go into this tool to use this thing and I'm going to go into this tool to use this thing.

Speaker A: Right.

Speaker B: I don't know about you Jay, but like I go into like a thing, ideally one thing and in that one thing I want to be able to do all the things and I don't want to have to be able to think about what it's doing. So, so can you like walk us through? Because that seems to be like, I was just thinking like the decentralized bit really was important with the, the kind of onslaught now with Gen AI and all the LLMs, because now that we can read those, we can now have this combined interaction with our data and, and how like in my mind like how advanced it was. You were already thinking about that because clearly, you know, you were designing Noe before ChatGPT and OpenAI. Like how, how were you thinking about that and how does that fit together in terms of like use cases for some of your, you know, best, I'll say applied uses of, of the tooling.

Speaker A: Yeah.

Speaker C: You know, so when you look at data, a enterprise, a company would have data that may be structured but also semi structured, unstructured data. Right. So ultimately the use cases are uh, how can that enterprise or how can a company drive value from that data quickly? So there are a couple of things uh, around that data side. It's the asking questions and getting, you know, your answers back is one thing, but you are getting the user to prompt that question. Right?

Speaker B: Yeah.

Speaker C: And that is, it's, it's really only about 20% kind of use case from a usage standpoint where the data really drives value. I think a lot of times is ah, that when it gets pushed to the underlying user with some actionable things that they can take or you take that and do something with it.

Speaker B: Yeah. Because it could be like overwhelming if you just push a lot of data. It's like, well, what am I going to do with this?

Speaker C: Yeah, exactly. But it could be as simple as, you know, from a workflow standpoint. I mean the way we run our company is across all our data within the company. We have Slack integration that Noe pushes it into. Right. So sales team knows their numbers every day on the opportunities for the week, uh, the tracking M for the month and the quarter and so on. The success team knows what the usage is looking like. Are there anomalous patterns on the Usage that needs to be acted upon and things like that. Right. So it's like. But that's proactively sent to them instead of them having to go look for it. And that makes a uh, fairly you

Speaker B: know, decent size in their workflow.

Speaker C: Yeah. But then there's the ad hoc use case is also important because you oftentimes you may. You know before I get into a meeting sometimes with a uh, with a potential customer, I need to know everything about them and I'll just go slash no. What's the. What was the next step with xyz? Right. And then it gives me an answer right from Slack. So it's a combination of all of that. So you can think of how all of these things playing out for the future is that uh, essentially it's an agent for all your company data. Right. And then in the world of agents in the future that will play with other agents to do things that you want and hopefully automatically take some actions on your behalf.

Speaker B: Yeah. Ah, and I'm already starting to look at decentralized agency AI too. You and I could have a follow up. We could do a panel about that with my friend that runs Instructs IO which is kind of a decentralized mess. Yeah. You mean Matt Harwin could have a good discussion about that. And I think that that's so important because I think what people aren't thinking of right now with static reporting is how is this going to meet that future? That future where AI agents kind of interact. I do believe that they will also be interacting in decentralized ways. Jay, just like. And so I think that, I mean I never really liked band aided old technology anyway even though it was modern ish. But I think now the foundation that you have to have to be future relevant is the word I would like to use is really different and I would love to get your thoughts on that. Like what do you think it takes to be future relevant today? And in particular when it comes to

Speaker C: analytics, you know, if you're a. The answer kind of depends on where you are in terms of a company right now. So because oftentimes it's uh, you got three things. You got people, processes and tools that, that are bundled up together in teams. Right. So and we are at this interesting conjecture where it's. There is a revolution happening in some ways especially with AI. Right. So. So if you are, if you're a latest, you know, a larger company for me is like the best bet for you is it's going to be hard to change up your existing Workflows and stuff.

Speaker B: But you were already on that plane and dropped in another destination.

Speaker C: Exactly. So you're going to get left behind unless you basically adjust. And the best way to adjust is partner up with companies that are doing frontier stuff. Right. Because things are changing and so your existing processes and tools and so on may not be quite relevant.

Speaker B: You know, many years down, um, we actually see this. Like, that's a great example. We see it in insurance, where I'll call it risk modeling. You have your traditional risk modeling company, and then the company that owns that, which is the biggest risk modeling company in the world, invests at the A and B level of, uh, basically what would seemingly be their competitor. And I also see it in Microsoft, which is, you know, SQL is one of the biggest database structures in the world. Right. Microsoft's investing in like, actually one. One of the technologies I'm really, I'd say, like, I'm really high on right now, which is space and time, which is a decentralized HTAP database. And you and I are going to talk about that when we get to persistence of data. Microsoft is, you know, interested in technologies like that. And so I think what we do, Jay, is like, if we actually look at the biggest things in the world, we see that they're actually not isolated in terms of their technology, that they understand that you can't just rip out that steam engine because the track's already laid, but they're also investing in and fervoring the future of pioneering. And then they're probably trying to figure out how to. How to blend those two worlds together.

Speaker C: Exactly.

Speaker B: I guess, like that, like, kind of like it reminds me of the EV versus the gasoline automobile. And of course now we have hybrids. And so I think of it like that. And I think, like, I think that there's going to be a lot of hybrid. Talk to me, talk to our guests a little bit about. I think most people think of analytics, uh, as something that is served up to them. Talk to our guests a little bit more, if you don't mind clicking in. Because I just love the functionality of Noe when I get to talk to the data. Can you talk to them about what's behind that with the NLP and the AI? Like, can you. Can you maybe just like kind of drill down a little bit so they get an understanding of the technology behind being able to talk to your data?

Speaker C: Sure. You know, within that there's like two use cases, and they require different approaches. So the one use case is you have documents, policy documents. PDF documents and so on. Right. So that you can use, ah, a rag engine on it with AI to answer the questions that you want. Then the second bit a little bit more complex as a data side. Right. So you have data across multiple different sources. And then when you ask a question, the complexity is how does it know which, which data set and which database and so on where the data is stored. And if you have, say, something like, hey, um, how much sales do we do for this location since yesterday? And something simp like that for this location. Maybe a field in some table somewhere.

Speaker B: Right, that relational mapping.

Speaker C: Exactly, exactly right. So the AI engine isn't going to know what that means, you know, so then it becomes a. So when you ask that question, that question needs to. First of all, it needs to, though it's almost a vector search to say which data sets are most relevant for this question. And from that data set it needs to have the knowledge to say this location is a field in this table and so on. And then, so how do you add up all of the question I'm asking to that location since yesterday? Maybe a time field somewhere. So it basically does all of that and then puts the AI elements to basically just enhance the accuracy of it and send you those results. So where we're going with all of this is that when you ask a question that can span across policy documents across your semi structured data or structured data, have cohesive answers that give you across both of them, along with explanations on how it's come to the conclusion.

Speaker B: I love that. And I don't think people understand. I mean, I'm going to like give you props here, Jay. That's like as simple as if you're an Apple person and you have a thing called AirPods and you flip the case open and it uh, automatically says, would you like to connect? It can even do it on the airplane. Now that's as simple as that. So imagine for all my users out there, and I'm going to get to the executive version of this in a moment, Jay, because I, I want to talk to you about this executive version I had when I was like back in 2016 and you and I didn't know each other. Yeah, Imagine just being able to type or talk to your computer or your phone and say, I want to know about this and now I want to know about that. And you're, you're not talking to ChatGPT now, which I do and really enjoy talking. Me and my ChatGPT are besties. But now you're actually talking. That sounds A lot. But now you're actually talking to your data, your real data. And you're not just talking, Jay, to that one data set that's in that one database that's in that one thing. And you're not just talking to what happens to be in that data warehouse. To your point, that was structured. What about the semi structured? What about the unstructured? And you're not just talking to what some IT person coded it to talk too. Because on demand you might need, to your point about the vector logic, you may need to go beyond that. It's now able to, based on that conversation, to actually get to the right place and extract that information and give it back to you in a format that's not just hieroglyphics but is visual.

Speaker C: That's right. Yeah.

Speaker B: So impressive. Right? Like if we really think about that, that is so powerful. I mean, right. Like, I mean that's huge. How have you seen people respond to that?

Speaker C: Yeah, yeah, I do think so too. Yeah. Yeah. It's, you know, it's. You know, it's funny a uh, couple of years ago it, when you, when we gave demos and things to people, they were impressed by it, but they almost were a little intimidated to even put in the question. So the, but the last year and a half has changed. That is, I don't know how do I ask this thing, you know, so. And now even the expectations of what you can ask and so on change too. And yeah, and then I think what sometimes people fail to realize is that you can't chuck a whole database into an LLM and then have it expect because you're just going to run into context lanes and so on. Right. So it just requires different codes.

Speaker B: Chat GPT is like, uh, I've lost you. I mean even though it has in mem. Yeah, you're right. I mean there's only so much you can put in there.

Speaker C: Right, exactly. Right. So it just needs a different. It's a different kind of solution. But that merges up the, you know, the, as you said, the rails. You want the rails and you just want, you know, hyper fast train on it in some ways. Right. So instead of your Steam engine.

Speaker B: I love it. Okay, so now I'm going to give you my like executive wish that I had least. Let's go back in time, back in time to 2016 when like all of our little echoes and dots and everything first started coming out and I was running and like operations. So I had responsibility for thousands and thousands like 3,000 carriers and I had to know what was going on with losses. Because claims was in my domain, I had to know what was going on with our reinsurance treaties, our loss ratios, our profitability.

Speaker A: Right.

Speaker B: Like, all these things. And so j. My. And I think I can't remember if I've asked you this before or not. So. So we'll take it in the hot seat. I literally looked at my, uh, team and I was like, okay, here's what I want. I want to wake up in the morning and you can imagine this with the wildfires right now going on in California and all the things. You can imagine how many people across all walks of life, all business executives, from food to logistics to insurance, everyone is trying to get data information, like, at their fingertips. Okay? So my. And I was always under high stress environments in my role. I never had to answer easy questions. I wanted to wake up in the morning and say, I want to know this. Please tell me whatever it is, insert. I want to know this, this, and this. And I wanted to talk to Alexa because, uh, I tend to like, uh, and I was like, but I don't just want to know it. I want to based on you. Tell me the top five, tell me the top ten. Give me the exposure here, tell me this. What were my losses? Whatever it is, drop that into a report, send that report to my boss, schedule a meeting with her, and tell her I'll meet her at 9am when I get into the office. And by the way, then I wanted to go get my shower and get dressed for work now, Jay, when you're probably laughing and no one can see your face other than me, they literally thought I had, like, five heads. When I, when I asked, like, can you imagine, like, I literally asked out, uh, of my technical team back in 2016. They thought I was absurd, insane, and mental.

Speaker C: Yeah, I had a similar reaction. I mean, my team gave me a similar reaction too, back in 2017 and said, I want to be able to talk to Alexa and give me, give me the results back, you know, and go, oh, uh, yeah, Jay's off his head again.

Speaker B: So, so, so you and I were like, thinking. But you know what, here's what I. Okay, so now it's 2025, and now we have, like, not only the vision, but we have the underlying guardrails and frameworks. Now let's take that example, Jay, and tell me how we could realize that for insurance distribution. So if I want to know my top markets, my top carriers, my top place submissions today, how can we do that now? Is that realizable? And if so, how.

Speaker C: Yeah, I think, uh, it's, uh, I would say it's partly realizable. So you could get those top, top five things you're looking for and then have that sent over to you and so on. What would. There are a couple of things that we're expecting the market to do in time. And part of it goes back to the agentic approach. Right. So it's like if you're speaking into something or just having a conversation or interface to say, hey, what are my top five campaigns? Right. That that's worked and it gives you that info and then basically, hey, set this up as a report, send it to xyz. All of that stuff, uh, most of it is there in our platform. Some of it we are working on, especially on the actions piece.

Speaker B: I mean, I always expect to be people to be pushing that. Like.

Speaker A: Right.

Speaker B: You and I are clearly not people who were like happy with precondition reports back in 2015.

Speaker C: Yeah. But when you, when you put the future State hat on, it's like, you know, I don't know how long this is going to take, but like just me as a person, I would expect a personalized, you know, AI agent for me, right. That knows everything about me. M can read my email and all of that. And even when it wants to basically like do something simple as, hey, I need to give me the best insurance option for me that can crawl up different insurance sites and so on, and knows my specific, you know, things, the data associated to me, right. I've got a couple of cars. One of them is hard to insure and who's going to insure that and so on. So. And it's uh, the future as we see it is multiple things working together in tandem. Right? So where, uh, we fit into the picture for us is like, hey, no, it could be that data. Anything that you want to know about the data from a company standpoint and basically having other things that work together with it to drive the actions that you want. Right. So for example, in your case, what's the question you're looking for? No, he answered that question and then send this to this person. No, we can do that and then set up a meeting. No, he probably is not going to do that. It's going to be another agent that does that. So it's not basically combined.

Speaker B: That would be a calendly agent or whatever.

Speaker C: Exactly.

Speaker B: We'll give Tope. He's from Atlanta, he's from my, from my hometown, so we'll give him that. But so let's let's just kind of emphasize in order to do so we need the agency AI and I think that's going to start coming. I think it will be the shiny ball this year and I think in 2026 we'll see it start to take form. Right. Like I think we'll start the people that are working on the edge are working on agenc AI. And to everyone who's in monolithic technology, because we're going to go to decentralized in just a minute. Jay to everyone who is in monolithic technology, the best agency AI in the world within a monolithic technology that doesn't talk to other technology is not going to get the job done. So I'm just going to. I love seeing some of the biggest. I won't name names today monolithic technology in the entire world doubling down on their agencic AI. And I'm like, but it's only in system. Okay. So that leads me really well into decentralized data. And I just want to go to. I mean I really focus a lot on decentralized architecture and I think that that's so important and it's kind of the, the backbone of how you really built and kind of vision where the world was going to go with noe. Can you just come back for a moment and just emphasize that to the listeners? Because I think not a lot of people fully understand what decentralized architecture is. And so if you could just maybe break that down for people.

Speaker C: Sure, yeah. So decentralized data systems for us is like, hey, data is going to be in its natural state wherever it needs to be in. Right. So it could be in services that can be accessible by APIs. It could be in relational data stores. That's inside the company. It could be third party services, documents and so on. So it's like leaving it where it's at without basically having a central repo to manage it all, essentially. Right. So but you still need that single pane of glass across your assets. So that's kind of.

Speaker B: No, he's going to be that single pane of glass.

Speaker C: That's right, yeah. Um, that's uh, that's been objective.

Speaker A: Yeah, yeah.

Speaker B: And I think that that's really important because then that single plane of glass then has the AI and the querying capability on top of it. And then to your point, it is not modern ish. It is pioneering, ready for then becoming able to interact with other AI agents. So it's a single pane of glass that's like ready to do that AI AGENC AI handoff And to me that's what I look for when I think of on the edge processing and tools that I end up working with and thinking about are. Yeah, I mean we'll just call it like, you have to be like on the edge ready even if not all the other dots are connected yet. Even if the other agent on the other side isn't there yet. I was always thinking that you got to think that way, especially when it comes to reporting. Okay, so tell me Jay, from your perspective, what do you think, because we've talked a lot about like where we are, why you built it, what the future holds, what do you think is the most interesting thing that we will see emerge, you know, in the next, I'll call it 12 to 24 months. What do you really think is going

Speaker A: to be the future?

Speaker C: 12 to 24 months, that's fairly short term. You know, I think, I suspect a couple of things will happen. You know, the LLM side of things was that uh, you know, there'll be a plateau around that and hopefully someone figures out the, on the research side, a next level up. Right. Because right now people are focused on that token prediction for the next keyword and that's how these LLM engines are built. Right. So now the next step is merging that up with workflow specific use cases or domain specific use cases to drive value. Because instead of going to uh, the chatgpt or perplexities of the world and asking questions and giving you cohesive answers back is how can you take that and make your enterprise workflows easier? So we will still be I think in the beginning stages of that in the next 12 to 24 months. I mean oftentimes we are too optimistic um, on these things. But it's the nature of the beast. It'll be a little slower than that we might think. So yeah, I think it'll be the beginning stages of that.

Speaker B: All right. I think that, I think agency and I think you're right, I think we're going to see a, ah, normalization this year and then I think we're going to see, I think if I said it differently, if you and I look back at this recording like a year from now for people that didn't start now on securing and procuring that handoff capability and enablement and at least adding these pioneering tooling to their tool set and their arsenal, you are going to be behind a year from now because it's going to have, it's like when we all started first doing it, it'll sound stupid because I'm API less. But when we all first started doing APIs, Jay, and do you remember there were some people who weren't API enabled? And I'm like, okay, like no, you can't be here if you're not API enabled. And of course now we're like past APIs. APIs are actually in my mind really slow. And so now we're on real streaming. But okay. All right, so as we think about this, I think what's the action that you would do? I always lead with a call to action. What would you ask people to start doing, to stop doing and to continue to do? Because I think that's just such a great way to frame it.

Speaker C: Yeah, I think one thing that might be helpful for people and it was, uh, definitely be helpful for me and Noe as well, is just to in the context of what you're doing and the industry you're in and just to have a think about what that's going to be like in the next few years, uh, what things that you expect to happen. And then from there you can reverse engineer to say, hey, if I want to, if this is where things are headed, how do you get there bit by bit, right. And then you can take parts of that and then make things happen. And there may be small steps to making, uh, those happen. Either you know, the company you're in or even just personally too.

Speaker B: What do you mean? I'm just going to call you and say, jay, this thing that I just dreamed up, why don't you just have it? You're like, stay away from Lisa if you want to go bit by bit. No, I'm joking. I love it. I love it. Well, thank you. First of all, thank you so much for coming on Insurance Unplugged. I mean I, I love having guests that don't always eat, sleep and drink our industry 100% of their day to day and love having these really on edge discussions with emerging technology and things that people can do. Huge fan for everybody listening. If you don't know of Noe, it's K N o W I so definitely follow J follow his team at Noe and as I always like to say, stay curious, stay informed and we will definitely look forward to hearing your comments next time. Thank you so much for tuning in today on another episode of Insurance Unplugged.

Speaker A: Today's episode of Insurance Unplugged, the AI and distribution series is proudly sponsored by Iris Insurtech, your gateway to the future of insurance distribution. Iris harnesses the power of generative AI to transform data processing and decision making across the distribution landscape. The IRIS platform integrates AI driven decision engines, dynamic form generation and configurable workflows, all underpinned by continuous data quality management. Discover how IRIS is powering smarter operations and more efficient distribution. With cutting edge AI setting a new standard of excellence across the entire industry, it.