 Hi, this is your host Abhinav Bharti, and we are back with the Predictions video series. And today we have with us, Devani Lamas, CEO of Transposit. Devani, it's great to have you back on the show. It's great to see you again, Swapna. It's my pleasure to host you today. And today we are, of course, going to talk about Predictions. But before I ask you to grab your crystal ball and share your predictions, let's just quickly talk about what is the company all about? Transposit is a company that uses AI to help operations teams, so platform engineering teams, site reliability engineering teams, and engineers to better handle incidents, alerts, and kind of the whole process of being on call. Now it's time for you to pick your crystal ball and share your predictions with us. My first prediction is I think we're going to see the emergence of much smaller, niche focused AI models in 2024. I think right now everyone is spending a lot of time using what I call mega models, right? Like the big trend is how big can these models get? Right now open AI continues to be the leader in that market with GPT-4, which is powering tools like chat GPT. But I think you're going to start seeing the emergence of much more focused application and domain specific models for lots of reasons. But a couple of the biggest ones, just being the ability to get much more cost effectiveness out of those models, as well as to increase the accuracy of those predictions, which is going to be really, really critical as people start rolling out AI into production workloads in the coming year. My second prediction is that 2024 is the year that AI will move beyond the chat interface. I think we've all used different tools that are leveraging chat. Chat GPT is the big one right now, but lots of people have been rolling out tools, Google's barred and so on that are kind of sitting there and like living next to your interface with much more of a chat interface. However, I think that what we're going to start seeing is really AI going mainstream into business workflows. So what do I mean by that? I really mean you're going to start using tools that you don't know are necessarily leveraging AI behind the scenes to help with correlation, help with data analysis, data processing. But more importantly, I think you're going to just start seeing AI fitting into the applications that we use on a day-to-day basis where you don't necessarily require that back and forth of chat, but instead it's just a seamless part of the experience that helps us use our tools better. And my third prediction is a little bit controversial. So it's a little bit controversial of a prediction, but right now I think a lot of people are talking about the RAG stack as one of the pieces that they're most excited in AI. And for anyone who hasn't been keeping up with every week by week on what's going on with AI, the RAG stack is retrieval augmented generation. So what that means is using AI to access external sources of data, knowledge bases and other things like that. However, I think that right now most of the examples of that RAG stack are very basic and are really primarily used for keyword search. And so I think that what you're going to start seeing is that LLMs are the best tool for ETL that has ever been created and you're going to start seeing much more sophisticated pipelines for processing data, for adding application context, for doing classification and categorization in domain specific areas prior to ingestion, which I think is going to finally really start delivering on the promise of AI as a really fabulous cognitive engine for helping us correlate across all of our different data sources. Excellent, thanks for sharing those predictions. What kind of challenges you see will be there in 2024, not only just for customers, user, larger ecosystem, but maybe also for transposite to tackle. I think that the big challenges that I'm hearing from our customers right now are, they're twofold, right? One of them is there's a lot of noise in the market and a lot of people talking about AI and how it's going to fit into their workflows. And I think that a lot of people are kind of like, as a vendor, we're challenged to showcase how our approach is different than other people's. And I think for buyers, the really big question is, how do you decide between different tools that are now all promoting AI for you to really figure out the ones that are going to drive a business advantage for you? And the second really big challenge, I think is an evergreen one, but one that's becoming even more important in 2024, which is, there's a lot of data out there that we want to leverage in our tooling and AI means that we can take advantage of that data in a way that we haven't been able to in the past. However, a lot of that data is still very messy, still very unstructured, hasn't necessarily been cleaned up. And so I think people are asking the question of how do they prepare their data, their workforces, their application stacks to really be able to take advantage of kind of all of the wonderful things that are coming down the pipe in the AI ecosystem. Excellent, thank you. And what is going to be the focus of a transposite in 2024? Our big focus right now is on unstructured data. So what I mean by unstructured data is human-generated data. So that's to say the Slack conversations, the transcripts from video calls and interviews, you know, kind of the postmortems from incidents and really helping our customers take advantage of this very underutilized and very messy data set that they haven't really had, you know, observability into in the past. And specifically in our world, in the context of on-call alert and incident management. So, you know, kind of really, really working on taking this new data set and using the power and magic, honestly, of AI to make it powerful and usable for people so that when they are solving issues, you know, we can stop relying so much on institutional knowledge and really start leveraging AI for what I think it is, which is a force multiplier for our teams. Devani, thank you so much for taking time out today. Share your predictions. Of course, I would love to have you back on the show next year to see, first of all, how many of these three predictions turn out to be true? And then get next set of predictions for the next year, but I really appreciate your time today. Thank you. Thank you, Swapna. I'm really looking forward to it.