 So thank you everyone. Thanks for being here through the day. I understand you've gone through a lot of content since morning. Um, and everybody's talking about new age technology, how you can use technology for your marketing campaigns. And what's technology without AI today, right? If we don't talk about AI, probably there is no conference that goes ahead without speaking about AI. Um, before we move further into my presentation, I just want to spend a minute talking about what we do, right? Probably some of you would have heard what double verify does, but double verify as introduction says is a guardian of media quality. And what I mean by matter of media quality is we measure where you're buying impressions where you're buying inventory weather is going in a brand safe environment, whether the placements are done in a content which resonates your brand ethos, whether the impression you're buying is a human impression or is a fraudulent impression, right? Whether the impression you're buying is delivering the viewability or it's not even delivering the viewability or the opportunity to be seen by a real human, right? And whether it's delivered in the right geo, so brand safety, ad fraud, viewability in geo and obviously optimizing your spends to meet the KPI and deliver efficiencies in your campaign, right? So that's what we do. So next my 10, 10, 15 slides that I have will mainly touch upon AI in these realms and how we as an organization using AI to make it more relevant for our customers and clients so that they bring in efficiencies in their media plans, end up saving money, reach to right audience in the right context and in the right environment, right? That's where we are. So that's me, I lead the India business for double verify as a sales. Broadly, there are there are two agenda points which we touch upon, right? Perils and profits in the age of AI and obviously the opportunity for us that AI brings a to increase or or better the media quality and be bringing in more efficiencies in your media that that you buy. So we'll first start with perils and profits in the age of AI. We're talking about how consumers are changing, right? We all know that there is no longer a linear journey of a consumer, right? It's complex at one point of time you're reading a review at another point of time you're going on on e-commerce site to buy stuff. Other points are going to OTT. So earlier there was an IDA model that one used to follow, but there is no IDA model anymore, right? Every single user is multitasking consuming different set of content. And what are they doing, right? So about 74% Indians buy more items online today, right? Compared to couple of years back or pre pandemic stage, right? That's some massive, massive change. And how are they buying it, right? It's not, it's not only about their searching at one place going to e-commerce buying it, right? The search behavior has also changed. It's no longer only the search engines. It's the social media. It's the recommendation from your friends and family. It's the reviews that you see on social media. So there are multiple touch points for them to look at the information. So about 52% Indians use social media as a source of news too, right? And we know how it can be manipulated using AI. So we'll touch upon that soon. That's a massive number. 47% of Indians look for information about brands and products on social media. It's like no brainer, right? And very interestingly, from entertainment perspective, right? About 74% Indians subscribe to one more streaming service in past two years. That's massive, right? So if there were 10 people who are looking at one OTT in the year back, there are 20 people who are looking at two OTTs now, right? So number has literally doubled. So what it has done, it has kind of fueled content generation, content consumption, which also puts brands at risk because you never know where the content is coming from. You never know where the ad is going to go and whether that placement of my ad is in a right environment and in front of the right audience or not, right? So it kind of brings its own risks. Now before we move ahead, right? From AI perspective, I just keep the floor open for 10 seconds. Anyone of you can just shout out. In your perspective, in your organization, how you're leveraging AI? Anybody, I think it's not a right idea to throw open question in front of a large audience, but anybody who wants to give a shout to speak about, are you aware of how AI has been used in your respective organizations or your respective organizations. Anybody, any random answer is fine. Just 10 seconds I have. No. So let me, let me put it there, right? So there are multiple ways in which you can use AI and essentially the baseline for using AI is bringing in efficiencies across different work streams, right? So to begin with operational efficiencies by automating your repetitive tasks. So it can be generating your frequently required reports or it can be creating insertion orders or POs and stuff like that. So AI is bringing in more efficiency there. Smarter strategic planning, faster decision making. So AI is literally now doing your base level analysis where you used to need one analyst to do that. AI is doing it at one go at a scale and returning the outputs probably in few seconds. Impactful and compelling ads. Generative AI is the buzzword, right? Everyone knows Generative AI. Now, what you need, the 10 operators, what you should do now is being done by AI in a minute's time. And of course, from campaign perspective, meaningful insights to drive campaign effectiveness. We spoke about insights and analytics, right? Now it's also creating challenges for marketers in different spheres. Content quality, fraud, scale and complexity. And I'll talk about each of it. How? So if you look at what used to happen few years back, couple of years back, not even few years, couple of years back is probably they're aware about dozens of articles on a website, right? And it's to take certain time to create a piece of article and upload it on your website. So there were limitations to create content there. With the help of AI, now you can literally create thousands of articles in a span of a minute and put that as a piece of content on a destination. Look at the scale. If you have hundreds of thousands of websites bringing in 1000 pieces of 10,000 pieces of content in short span of time, imagine where the world is going, right? Estimates are about 99% of digital content will be AI driven by the end of 2030. And that's massive. And it's almost 100% of the content will be AI driven. And about 19% of programmatic bit streams are going to be AI powered. And why we say that? Because once this scale of content is generated, the AI will have to play a role to bring in more transparency in the programmatic ecosystem as well. Because that's where the supply will open up significantly. The demand probably will not be as much. And obviously the fraudsters will start playing games around it. And as we're talking about fraud, right? So fraud follows money and AI, the fraudsters are also leveraging AI big time. And it's across everywhere. Mobile app schemes where we've seen two times more fraud signatures coming in, under 37% more ad fraud schemes coming in on audio devices. And CTV obviously is a buzzword. And we see a lot of fraud and activities happening on CTV inventory as well. So about 15% of your CTV impressions are served when the screens are literally off, right? And we are not aware of these facts. It's Reels, the era of Reels. It's the era of shots. Obviously, TikTok is not in India anymore. But yes, this is the era of the youth who is a Gram generation as everybody calls it. And it's UGC content. UGC content can become really, really difficult to analyze and becomes really difficult for brands to stay away from. Because it's about 160 million hours of video viewed per minute only on TikTok, right? Probably we see similar behavior on Reels and shots in India if you draw parallels. And there is complexity of content, right? There is text, there is video, there is audio. How you as a brand will know what content is good for me, what content is bad for me, right? So these are the challenges which AI is throwing in front of us as a marketer. And that's where, if there are bad elements were using AI, then probably as guardians we also need to take help of AI to kind of counter that program, right? So how we as an organization are capturing this opportunity. So using AI to detect fraud at scale. How we're doing it? We'll explain that. So machine learning models paired with human insights is something which is helping us to detect fraud at scale, report fraud at scale and help advertisers save their media money. Classification, when I say classification, it's how we categorize different pieces of content in different categories, right? So there is a piece of content which talks about, for example, alcohol, right? Now alcohol can be a cocktail recipe or alcohol can also talk about drink and drive issue, right? So how that AI really captures the context of it and puts in a bucket which is about cocktail recipes or about vehicle disaster. Right? And how we can help our customers and brands today to safeguard their investments and not show their ads in vehicle disaster category, but they're okay with the recipe category, right? And obviously predictive AI, which is the next big thing which helps us optimize the campaigns at scale and bring in more efficiencies, right? So if you're buying something at X cost today, how you can save probably X percentage of that cost by optimizing it without any manual intervention and meet the KPIs that you've already said, right? I'll touch upon it. I'm just running out of time, but I'll touch upon it quickly. So from classification perspective, how we classify a video, right? It's slightly hazy image, but if you see how AI reads this as a video. So AI sees there is a wine bottle out there, right? AI sees there is a glass. AI sees there is a small handbag lying on the table. AI sees it's a table set. So probably what it will do is it will classify this video in alcohol low risk, which is a pink box that you see. If it would have been a car, it would have been an accident, then this classification would have been alcohol high risk because it talks about vehicle disaster or an accident where no brand wants to be seen on, right? So you can simply remove that from your campaigns. What's the sentiment? The sentiment is positive. So it will also categorize that video as positive sentiment. It will also put it into contextual categories like travel or food and drink or luxury and things like that, right? And then obviously as a marketer, you can pick and choose whether you want to be or your ad can be shown in this kind of content or can't be shown in this kind of content. You can literally block that piece, right? So that's how the classification has been done by the AI at scale, yeah? And how do we kind of, you know, activate AI for dynamic optimization, right? So there are multiple inputs that goes in the AI system, right? So we call it AI as a dragon, right? You train it the way you want. And the more you train it, the more it starts following you, listening to you and doing things the way you want it to do, right? So there are multiple inputs today from campaign level that you can give to AI, right? So there can be something like as you can set campaign objectives as reduce my ECPM or you can say that maximize my convergence or increase variability to a certain stage, right? That goes as an input from K-Pay perspective. Then there is cost data that goes in, right? What bid you want it, maximum bid that you want to set it up for? There is measurement data from brand safety perspective, variability perspective, NGO, fraud, all that data kind of goes in this AI system, right? There are first party data that also can be fitted in and there is DSP data, right? Now AI gets all this data starts analyzing on more than probably 200,000 parameters for which it can bid on. And it bids on the best quality impression basis the KPI that you have, right? And it starts dynamically optimized. Now what used to happen so far is you have a set of traders or agency experts who are sitting on a desk looking at the campaign and they start optimizing the campaign on their own, right? But it has its limitation. They can't do it real time. They can't do it probably every hour. AI works continuously, looks at more than 20000 parameters and then bids the best quality inventory that you deserve and also pay probably lesser and give the fair price. I wouldn't say pay lesser, but give a fair price for that transaction. Which is what gives you outputs for you to measure on and that output again is works as an input for the AI to start optimizing it on further, right? Slightly complex, but from AI's perspective it kind of works to bring in more efficiencies and what efficiency drives, right? Media efficiency, right? Maximize campaign reach on quality media inventory by optimizing your quality CPM, right? We are not talking about CPMs anymore. We are talking about quality CPMs now, right? Because as a marketer, if you are paying X money for an impression that you are buying, you get 100% of the quality impressions in return of that money. There is no compromise on that. There is 0% tolerance there, right? Increase return on ad spends, right? Everybody, every CMO has this as a KPI probably on top of their minds, right? How do I increase return on my ad spends? AI kind of helps you increase that return by focusing on impressions that work and give fair price for that transaction, right? And obviously, attention measurement is a buzzword now, right? So how do I measure attention and go beyond just standard measurable viewability now? And how do I optimize on attention? So in real time, AI now starts optimizing on attention, which again works to deliver better return on ad investment for customers. Simple one case study I have. You can't name the brand. It's an automotive brand which using AI, how did we bring in more efficiencies there? 124% increase in attention index, 340% CTR improvement, 7% less CPM. And obviously, if this data can be fed into your brand list studies and stuff, you will get more enriched insights basis data. Yes, that's it from me. This was my last slide. So what we do is not only just keeping your investment safe, but drive outcomes by building in efficiencies and performance in your media plans. This was the only sales slide that I have. But thank you so much for listening to me patiently. I hope it was useful. I'm available backstage if you want to talk. Thank you so much. Thank you. Thank you so much, Nachik.