WEBVTT 00:00:00.542 --> 00:00:01.334 Nevash, welcome. 00:00:01.918 --> 00:00:03.878 What are you seeing and what have you seen, I suppose, 00:00:03.878 --> 00:00:08.717 over the last two years of the impact of GenAI on customer experience? 00:00:09.050 --> 00:00:12.345 So the first thing is, I'd say over the last two years, 00:00:12.345 --> 00:00:14.806 there's been a fair few proof of concepts. 00:00:16.182 --> 00:00:20.770 Telecoms in particular get that with millions of subscribers, 00:00:20.770 --> 00:00:24.190 hundreds of products, thousands of employees, 00:00:24.190 --> 00:00:30.238 you need help with AI to actually get to the problem faster and solve the issue. 00:00:30.238 --> 00:00:36.369 The change to me has been the volume and variety of data that they have access to 00:00:37.454 --> 00:00:39.414 that has accelerated that. 00:00:40.582 --> 00:00:44.836 You know, from a Databricks perspective, the best use cases that we see are 00:00:44.836 --> 00:00:49.299 actually in production and they're in production because there are three things 00:00:49.299 --> 00:00:50.383 that are happening. 00:00:51.259 --> 00:00:54.637 The first thing is the data has to be the right clean, 00:00:54.637 --> 00:00:58.058 organized data for GenAI to actually work effectively. 00:00:58.475 --> 00:01:01.603 So if you have siloed data in your organization, 00:01:01.603 --> 00:01:04.939 the GenAI is going to represent one of those silos. 00:01:05.148 --> 00:01:09.611 But where you have the biggest impact is where the data is together because then 00:01:09.611 --> 00:01:11.571 you have a 360 view on the customer. 00:01:12.072 --> 00:01:13.990 The second area is having governance. 00:01:14.699 --> 00:01:20.580 GenAI works best when you can tell who's accessing the data, how it's transforming, 00:01:20.580 --> 00:01:23.041 and how it's delivering an outcome. 00:01:23.333 --> 00:01:25.085 And then of course, you have your AI on the top. 00:01:25.085 --> 00:01:29.714 And it's not just about having the model, it's really about governing the model, 00:01:29.714 --> 00:01:31.674 evaluating the model, changing it. 00:01:32.092 --> 00:01:35.345 And the best use cases I have seen relate to growth, 00:01:35.345 --> 00:01:41.810 growing the business while delivering the best customer experiences and driving efficiency. 00:01:41.810 --> 00:01:44.354 I want to talk a bit more about that evaluation, 00:01:44.354 --> 00:01:48.525 that measurement because it's one thing to be very excited about this new 00:01:48.525 --> 00:01:51.194 technology and how it can benefit your company. 00:01:51.194 --> 00:01:53.696 But the other is how do you measure it? 00:01:53.696 --> 00:01:54.989 How do you measure it's impact? 00:01:55.573 --> 00:01:56.366 Are we getting it right? 00:01:57.992 --> 00:02:01.704 Look, I think that is work in progress and it 00:02:01.704 --> 00:02:03.665 depends on the use case. 00:02:04.290 --> 00:02:09.254 So some of the use cases that have been evaluated or there's a value associated 00:02:09.254 --> 00:02:14.134 with it from our customers, you know, there's one that's just growing the SMB 00:02:14.134 --> 00:02:18.847 business and small to medium businesses is a key area of focus for telecoms 00:02:18.847 --> 00:02:22.642 because, you know, generally the CIOs make decisions faster. 00:02:23.059 --> 00:02:26.688 So we have a customer who's already grown by 20%. 00:02:27.230 --> 00:02:31.109 And what the AI is doing is what a human cannot do very quickly. 00:02:31.526 --> 00:02:34.445 It's not just looking at one data source before they call the customer. 00:02:34.445 --> 00:02:35.655 They're looking at complaints. 00:02:35.655 --> 00:02:37.866 They're looking at what the customers bought previously. 00:02:38.158 --> 00:02:41.244 They're looking at their capacity against the plan they have. 00:02:41.578 --> 00:02:46.207 And these multiple sources are resulting in a much richer conversation. 00:02:46.207 --> 00:02:49.627 And that gives you the ability to grow into more use cases. 00:02:50.086 --> 00:02:51.713 I know it's pretty early days. 00:02:51.880 --> 00:02:55.341 Yeah, well, I say, I say that, but actually AI has been around for a long time. 00:02:55.341 --> 00:02:58.553 But in terms of this new generation, I suppose, of technology, 00:02:58.553 --> 00:02:59.596 it seems early days. 00:02:59.804 --> 00:03:00.889 Where could this go? 00:03:01.139 --> 00:03:06.561 I spent 20 years in telecom myself and I have worked with more than 200 telcos 00:03:06.561 --> 00:03:08.271 over the last four years. 00:03:08.563 --> 00:03:13.151 So to me, where this could go, I mean, the possibilities are endless, 00:03:13.151 --> 00:03:17.363 but I see the biggest impact in growing your revenue, you know, 00:03:17.363 --> 00:03:19.240 better customer experiences. 00:03:19.240 --> 00:03:21.492 Now, mind you, I don't think customers actually care 00:03:21.492 --> 00:03:23.119 about the technology that anyone uses. 00:03:23.119 --> 00:03:27.415 They want the problem fixed before they even know. 00:03:27.415 --> 00:03:34.255 And I think that's what we have the ability to do with agentic AI and I think efficiency. 00:03:34.255 --> 00:03:40.178 But my most favorite use cases are the ones that impact the world in the best 00:03:40.178 --> 00:03:46.226 way, you know, make innovation happen, help in the health industry, as well as, 00:03:46.226 --> 00:03:49.437 you know, creating better business models. 00:03:49.437 --> 00:03:51.105 I think it's a win-win for everybody. 00:03:51.105 --> 00:03:53.691 And yeah, I'm excited about being part of that journey. 00:03:53.691 --> 00:03:55.777 Fantastic, Navash. Thank you so much for your time. 00:03:55.777 --> 00:03:56.486 Thank you.