WEBVTT 00:00:00.000 --> 00:00:03.600 Welcome to the Deep Dive. Today we are putting 00:00:03.600 --> 00:00:05.839 on our hard hats and getting tactical. We're 00:00:05.839 --> 00:00:07.740 trying to cut through all that vendor noise around 00:00:07.740 --> 00:00:11.039 big concepts like AI, like digital transformation, 00:00:11.160 --> 00:00:12.960 and go straight to the people on the ground. 00:00:13.080 --> 00:00:15.339 The CIOs, yeah. The ones who are actually signing 00:00:15.339 --> 00:00:18.019 the checks and delivering the results. Exactly. 00:00:18.199 --> 00:00:21.140 And we have this incredible stack of insights 00:00:21.140 --> 00:00:23.940 from a summit with some major enterprise tech 00:00:23.940 --> 00:00:26.359 leaders. That's right. I mean, these are CIOs 00:00:26.359 --> 00:00:29.670 who have steered companies. through huge growth, 00:00:29.870 --> 00:00:33.229 IPOs, major industry disruption. Some of them 00:00:33.229 --> 00:00:36.670 have been in that top tech seat. for, what, 10 00:00:36.670 --> 00:00:38.869 years or more? So our mission today is pretty 00:00:38.869 --> 00:00:40.590 simple. We want to skip the high -level theory 00:00:40.590 --> 00:00:44.030 and get right to the practical battle -tested 00:00:44.030 --> 00:00:46.329 wisdom. The blueprint. The blueprint, yeah, for 00:00:46.329 --> 00:00:49.030 how you actually implement this stuff when failure 00:00:49.030 --> 00:00:50.909 is not an option. We're going to be looking at 00:00:50.909 --> 00:00:53.710 voices like Elvina Antar, former CIO at Zora 00:00:53.710 --> 00:00:56.689 and Okta. And Mira Rajevel, the CIO of Palo Alto 00:00:56.689 --> 00:00:59.810 Networks. Plus Sharon Mandel from Juniper Networks. 00:00:59.990 --> 00:01:02.490 And what really stands out right away is just 00:01:02.490 --> 00:01:05.209 the sheer depth of their experience. I mean, 00:01:05.209 --> 00:01:09.170 you take Elvina. She spent six years at Zora. 00:01:09.469 --> 00:01:11.269 Right. Building the tech for the subscription 00:01:11.269 --> 00:01:14.310 economy. Exactly. And then she moves on to lead 00:01:14.310 --> 00:01:18.689 IT at Okta. And Mira, she came at it from a totally 00:01:18.689 --> 00:01:22.170 different angle, product engineering architecture. 00:01:22.230 --> 00:01:24.909 Her focus was cybersecurity. For over a decade. 00:01:24.930 --> 00:01:27.810 Yeah. Yeah. She led IT for three different, pure 00:01:27.810 --> 00:01:29.909 security companies. And then you've got Sharon 00:01:29.909 --> 00:01:34.170 Mandel. a serial CTO and CIO. She came from the 00:01:34.170 --> 00:01:36.989 media industry, which I find fascinating. Newstapers, 00:01:37.189 --> 00:01:39.810 television, that high -speed environment. Yeah. 00:01:39.950 --> 00:01:42.189 And she brought that adaptability into enterprise 00:01:42.189 --> 00:01:44.930 tech. It's also mentioned she took intentional 00:01:44.930 --> 00:01:46.950 breaks in her career for her daughter's life. 00:01:47.129 --> 00:01:49.030 Which I think says a lot. These aren't just technologists. 00:01:49.189 --> 00:01:52.469 They are seasoned business strategists who are 00:01:52.469 --> 00:01:54.629 playing the long game. OK, so let's unpack that. 00:01:54.709 --> 00:01:57.269 Before we dive into the big strategy around gen 00:01:57.269 --> 00:01:59.370 AI, I want to stick with those personal details 00:01:59.370 --> 00:02:01.930 for a second. Because I think they reveal a person's 00:02:01.930 --> 00:02:04.969 real philosophy. What are the gadgets these leaders 00:02:04.969 --> 00:02:07.290 are using when they are not running the enterprise? 00:02:07.590 --> 00:02:09.949 It's a great question because it really separates 00:02:09.949 --> 00:02:13.550 the professional from the personal. So Alvina, 00:02:13.569 --> 00:02:16.009 who builds these massive, scalable enterprise 00:02:16.009 --> 00:02:18.590 systems for a living, she uses the Oculus. The 00:02:18.590 --> 00:02:21.610 Oculus. For what, fitness routines? Not at all. 00:02:21.770 --> 00:02:25.569 Pure fun. VR poker. and get this, VR karaoke. 00:02:25.770 --> 00:02:28.289 VR karaoke, I love that. That is a commitment 00:02:28.289 --> 00:02:30.509 to fun, a total disconnect. A full disconnect 00:02:30.509 --> 00:02:32.870 strategy, yeah. Yeah. And then you contrast that 00:02:32.870 --> 00:02:36.889 with Mira Rajaval. Her professional life is intense 00:02:36.889 --> 00:02:38.949 cybersecurity, right? All data. So her personal 00:02:38.949 --> 00:02:41.150 tech is probably similar. Exactly. Reflects that 00:02:41.150 --> 00:02:43.229 same data -driven discipline. She uses a Garmin 00:02:43.229 --> 00:02:46.270 and this whole integrated software system to 00:02:46.270 --> 00:02:49.090 meticulously track everything. Her wellness, 00:02:49.310 --> 00:02:51.729 her eating, her sleep, all while she's training 00:02:51.729 --> 00:02:54.830 for a marathon. For her, the gadget isn't the 00:02:54.830 --> 00:02:57.849 hardware, it's the data layer. That sounds incredibly 00:02:57.849 --> 00:02:59.689 disciplined. It's like applying professional 00:02:59.689 --> 00:03:01.629 rigor to your personal life. What about Sharon 00:03:01.629 --> 00:03:05.750 Mendel? Sharon is a different philosophy entirely. 00:03:06.270 --> 00:03:08.270 A deliberate retreat from the digital world. 00:03:08.449 --> 00:03:10.569 So she prefers to be away from tech? Completely. 00:03:10.990 --> 00:03:14.389 She spends time baking or doing ballet. Her favorite 00:03:14.389 --> 00:03:17.520 gadgets are Either a thermapen for cooking. Very 00:03:17.520 --> 00:03:20.500 tangible, measurable results. Exactly. Or her 00:03:20.500 --> 00:03:23.439 Alexa. But only for these super -specific things 00:03:23.439 --> 00:03:27.139 like IoT controls in her backyard. She mentioned 00:03:27.139 --> 00:03:29.120 a custom setting for lights and sprinklers they 00:03:29.120 --> 00:03:31.900 call the Fountain Run. The Fountain Run. I like 00:03:31.900 --> 00:03:34.740 that. So we have total digital escape, intense 00:03:34.740 --> 00:03:36.800 data tracking, and then micro -automation. Right. 00:03:37.039 --> 00:03:39.120 The thread here, I think, is discipline. They're 00:03:39.120 --> 00:03:41.780 all optimizing systems. And that brings us right 00:03:41.780 --> 00:03:45.000 to the biggest challenge. AI. It's a perfect 00:03:45.000 --> 00:03:47.840 segue. Alvina Antar kicks off the conversation 00:03:47.840 --> 00:03:50.639 with this incredibly high bar. She basically 00:03:50.639 --> 00:03:54.159 says AI is eating software and ultimately AI 00:03:54.159 --> 00:03:56.759 will save the world. Whoa, that is a massive 00:03:56.759 --> 00:03:58.780 claim. Is that just, you know, investor talk 00:03:58.780 --> 00:04:00.840 or does she really see it as a humanitarian tool? 00:04:00.919 --> 00:04:02.740 No, she backs it up. She sees it as the single 00:04:02.740 --> 00:04:04.719 biggest breakthrough for scientific advancement. 00:04:04.900 --> 00:04:07.900 Like new medicines. New medicines, material science, 00:04:08.340 --> 00:04:10.840 tackling climate change models. The scale of 00:04:10.840 --> 00:04:13.400 the opportunity, in her view, is just genuinely 00:04:13.400 --> 00:04:15.740 enormous. OK, but how do you get from saving 00:04:15.740 --> 00:04:19.079 the planet to running a quarterly budget? How 00:04:19.079 --> 00:04:21.800 does that translate to enterprise strategy? She 00:04:21.800 --> 00:04:24.899 grounds it in productivity, generating real, 00:04:25.279 --> 00:04:27.959 profound, and meaningful outcomes. It's about 00:04:27.959 --> 00:04:31.120 working smarter. But the core challenge she points 00:04:31.120 --> 00:04:34.259 to isn't the tech itself. It's the people. It's 00:04:34.259 --> 00:04:37.160 the expectations. The expectation shift that 00:04:37.160 --> 00:04:39.399 AI is creating. And this is so crucial for you 00:04:39.399 --> 00:04:42.019 to get. What does that mean, specifically, a 00:04:42.019 --> 00:04:44.519 shift in expectations? AI tools are creating 00:04:44.519 --> 00:04:47.899 a new normal for instant access, instant answers. 00:04:48.540 --> 00:04:51.000 She argues that self -service is now the absolute 00:04:51.000 --> 00:04:53.720 baseline. For customers and employees? Both. 00:04:53.920 --> 00:04:55.759 We are past the point of submitting a ticket 00:04:55.759 --> 00:04:58.040 and waiting two days. People now expect to help 00:04:58.040 --> 00:05:00.339 themselves right now with accurate information. 00:05:00.500 --> 00:05:03.459 Which means every single role, from HR to finance 00:05:03.459 --> 00:05:06.319 to IT, has to completely rethink how they deliver 00:05:06.319 --> 00:05:08.759 services. Exactly. The whole support model is 00:05:08.759 --> 00:05:10.920 fundamentally broken. The revolution isn't the 00:05:10.920 --> 00:05:13.600 code, it's the service delivery model. And that's 00:05:13.600 --> 00:05:16.420 why the blueprint from MiraRajaval is so important, 00:05:16.480 --> 00:05:19.040 because it tackles that expectation shift head 00:05:19.040 --> 00:05:21.829 on. It does. Let's get into it. So Palo Alto 00:05:21.829 --> 00:05:24.990 Networks was already a heavy user of, let's say, 00:05:25.410 --> 00:05:27.689 traditional machine learning. Oh, yeah. Thousands 00:05:27.689 --> 00:05:30.550 of ML models in their firewalls and their security 00:05:30.550 --> 00:05:33.250 tools. But generative AI was different. So how 00:05:33.250 --> 00:05:35.889 did they approach it as an organization? They 00:05:35.889 --> 00:05:38.290 knew they needed a new strategy, a hybrid one, 00:05:38.610 --> 00:05:41.189 and they moved incredibly fast. This was back 00:05:41.189 --> 00:05:43.930 in late 2022. They went with both a top -down 00:05:43.930 --> 00:05:46.910 mandate and a bottoms -up innovation drive. OK, 00:05:46.970 --> 00:05:49.129 let's start with top -down. What does that look 00:05:49.129 --> 00:05:50.689 like? I mean, it's more than just telling people 00:05:50.689 --> 00:05:53.790 to go use chat GPT, right? Way more. It was built 00:05:53.790 --> 00:05:56.779 on education and accountability. They immediately 00:05:56.779 --> 00:05:59.759 set up these monthly AI days for their top 60 00:05:59.759 --> 00:06:02.180 to 70 leaders. And how did they pick those leaders? 00:06:02.319 --> 00:06:04.540 This is key. It wasn't just by title. It was 00:06:04.540 --> 00:06:06.639 by their attitude, their ability to actually 00:06:06.639 --> 00:06:09.639 drive change. And they brought in outside experts, 00:06:10.139 --> 00:06:13.040 big names from IBM Watson, from Google, to educate 00:06:13.040 --> 00:06:15.240 the whole company on what was possible. So they're 00:06:15.240 --> 00:06:17.379 leveling up their leadership first. Exactly. 00:06:17.540 --> 00:06:20.839 And they set up crystal clear ownership. Mira, 00:06:21.220 --> 00:06:25.389 a CIO, was accountable for enterprise AI. So 00:06:25.389 --> 00:06:30.569 internal stuff, HR, finance, IT. And the product 00:06:30.569 --> 00:06:32.689 side. The chief product officer was accountable 00:06:32.689 --> 00:06:35.550 for product AI. Yeah. That clear split prevents 00:06:35.550 --> 00:06:37.850 the confusion you see in so many other companies. 00:06:38.029 --> 00:06:40.290 That makes a ton of sense. OK, now let's talk 00:06:40.290 --> 00:06:42.290 about Bottoms Up because this is where you can 00:06:42.290 --> 00:06:45.449 see real money driving innovation. Yes, the AI 00:06:45.449 --> 00:06:47.370 mastermind challenge. It's a great model for 00:06:47.370 --> 00:06:49.889 anyone listening. They didn't just ask for ideas. 00:06:50.050 --> 00:06:53.399 They put money behind it. Real money. $25 ,000 00:06:53.399 --> 00:06:56.339 for each of the top five ideas and $10 ,000 for 00:06:56.339 --> 00:06:59.000 the next 10. That is a significant carrot. But 00:06:59.000 --> 00:07:01.100 you have to wonder, doesn't that prove that innovation 00:07:01.100 --> 00:07:03.420 isn't just happening organically? Seems like 00:07:03.420 --> 00:07:05.759 an expensive way to get employees to do their 00:07:05.759 --> 00:07:08.420 job. It's a fair question, but from their view, 00:07:08.720 --> 00:07:11.019 the cost was totally justified by the speed and 00:07:11.019 --> 00:07:13.040 the scale of the results. How many applications 00:07:13.040 --> 00:07:16.180 did they get? Over 600 in four weeks. The challenge 00:07:16.180 --> 00:07:17.860 wasn't just about the money. It gave people a 00:07:17.860 --> 00:07:20.480 protected space and senior sponsorship to run 00:07:20.480 --> 00:07:23.800 with an idea. And the result was seven key initiatives, 00:07:24.339 --> 00:07:27.220 including four new product lines building out 00:07:27.220 --> 00:07:31.040 copilots. A staggering return. But the enterprise 00:07:31.040 --> 00:07:35.000 goal was even crazier. A 90 % ticket reduction 00:07:35.000 --> 00:07:38.779 across IT, finance, and HR. 90%. That sounds 00:07:38.779 --> 00:07:40.980 like marketing. How did they justify a number 00:07:40.980 --> 00:07:43.620 that big? Well, it was necessary. They're a 20 00:07:43.620 --> 00:07:45.439 ,000 person company, and they were submitting 00:07:45.439 --> 00:07:48.899 something like 480 ,000 internal tickets a year. 00:07:49.279 --> 00:07:52.930 Wow. That's almost 25 tickets per employee. Exactly. 00:07:53.089 --> 00:07:56.189 It's a huge drag on productivity. If AI couldn't 00:07:56.189 --> 00:07:58.889 fix that, then the investment just wasn't worth 00:07:58.889 --> 00:08:01.970 it. OK, so they set this massive goal. What did 00:08:01.970 --> 00:08:04.389 they actually learn about GenAI's limits when 00:08:04.389 --> 00:08:06.850 they tried to implement it? This, for me, is 00:08:06.850 --> 00:08:08.769 the biggest so what of this whole deep dive. 00:08:08.970 --> 00:08:11.310 They learned very quickly that while GenAI is 00:08:11.310 --> 00:08:13.470 amazing for retrieving and summarizing information, 00:08:13.790 --> 00:08:16.569 on its own, it could only directly solve about 00:08:16.569 --> 00:08:19.290 20 % of those tickets. Wait, hold on. Only 20%. 00:08:19.370 --> 00:08:21.970 If the tool can only solve one -fifth of the 00:08:21.970 --> 00:08:24.550 problem, how do you bridge that massive gap to 00:08:24.550 --> 00:08:26.970 get to 90? It sounds like the technology failed. 00:08:27.310 --> 00:08:29.170 Well, and that's the thing. It wasn't a failure 00:08:29.170 --> 00:08:31.589 of the technology. It was a failure of the process. 00:08:31.810 --> 00:08:35.250 The process. The process. Gen. AI can find information, 00:08:35.590 --> 00:08:37.769 but if that information is in three different 00:08:37.769 --> 00:08:41.490 silos, or the process for, say, a travel expense 00:08:41.490 --> 00:08:44.950 has five steps that shouldn't even exist, the 00:08:44.950 --> 00:08:47.769 AI just spits out the complexity faster. So to 00:08:47.769 --> 00:08:50.389 get to that 90%, they had to go back to old school 00:08:50.389 --> 00:08:53.250 IT discipline. Exactly. They had to couple Gen. 00:08:53.389 --> 00:08:57.029 AI with intense process simplification and automation. 00:08:57.289 --> 00:09:00.490 That other 70 % was closed by saying, OK, before 00:09:00.490 --> 00:09:02.909 the AI even touches this, let's cut out three 00:09:02.909 --> 00:09:05.490 useless steps, standardize the last two, and 00:09:05.490 --> 00:09:08.049 then automate the decision. So the AI becomes 00:09:08.049 --> 00:09:11.070 an amplifier for good processes, not a bandage 00:09:11.070 --> 00:09:13.289 for bad ones. You got it. Think about employee 00:09:13.289 --> 00:09:16.669 onboarding or a complex IT access request. The 00:09:16.669 --> 00:09:19.350 AI can summarize the existing documentation, 00:09:19.450 --> 00:09:21.909 but if the documentation is terrible, the AI 00:09:21.909 --> 00:09:24.230 is useless. So you fix the documentation, you 00:09:24.230 --> 00:09:26.629 automate the identity check, and then GenAI handles 00:09:26.629 --> 00:09:28.549 the last little bit. That's it. It's the new 00:09:28.549 --> 00:09:31.149 tech amplifying the impact of old school discipline. 00:09:31.389 --> 00:09:34.370 And the timeline is just astounding. What's the 00:09:34.370 --> 00:09:37.990 status on that 90 % goal? MIRA reported that 00:09:37.990 --> 00:09:40.669 within six months, they were already 50 % of 00:09:40.669 --> 00:09:42.549 the way there. They cut their internal friction 00:09:42.549 --> 00:09:45.950 in half. In six months. Half a year. just by 00:09:45.950 --> 00:09:50.070 combining the new tech with the frankly boring 00:09:50.070 --> 00:09:53.049 but necessary work of simplification. That kind 00:09:53.049 --> 00:09:55.809 of rapid change leads us to a really provocative 00:09:55.809 --> 00:09:58.309 question that the moderator, Wendy Pfeiffer, 00:09:58.850 --> 00:10:03.570 asked. Is digital transformation, is DT even 00:10:03.570 --> 00:10:05.990 a thing anymore? Right. Is it just an assumed 00:10:05.990 --> 00:10:08.490 state of being now? And Sharon Mandel had, I 00:10:08.490 --> 00:10:10.490 thought, the perfect answer. Yeah, she argued 00:10:10.490 --> 00:10:12.950 that DT is forever a thing, not a destination. 00:10:13.500 --> 00:10:16.039 She points out that 10 or 15 years ago, a lot 00:10:16.039 --> 00:10:17.860 of companies were still on paper. The bar was 00:10:17.860 --> 00:10:20.159 just lower. The bar was so much lower. But the 00:10:20.159 --> 00:10:22.539 real difference now is we recognize these technology 00:10:22.539 --> 00:10:24.500 waves are just coming faster and faster. You 00:10:24.500 --> 00:10:27.080 finish one, and the next one, cloud, security, 00:10:27.259 --> 00:10:29.960 now AI is already here. I love the analogy she 00:10:29.960 --> 00:10:32.360 used for this, the mountain climbing one. It's 00:10:32.360 --> 00:10:34.980 so good. She describes DT like climbing a mountain. 00:10:35.500 --> 00:10:37.460 Each transformation, like moving to the cloud, 00:10:37.620 --> 00:10:40.330 gets you to a new vista. A new peak. You stop, 00:10:40.490 --> 00:10:43.230 you celebrate. You look around. But the second 00:10:43.230 --> 00:10:46.250 you do that, you see another higher, more challenging 00:10:46.250 --> 00:10:48.330 peak up ahead. It's a never -ending journey. 00:10:48.549 --> 00:10:50.470 And the climbing requires more than just buying 00:10:50.470 --> 00:10:53.570 new software. It requires business reform. Absolutely. 00:10:53.710 --> 00:10:56.909 For her, DT is about breaking silos. It's about 00:10:56.909 --> 00:11:00.210 changing how IT works. The line between IT and 00:11:00.210 --> 00:11:03.129 the business has to become, as she says, less 00:11:03.129 --> 00:11:05.580 and less distinguishable. The technology is just 00:11:05.580 --> 00:11:07.659 the vehicle for the current climb. The business 00:11:07.659 --> 00:11:10.759 model reform is the forever thing. Exactly. So 00:11:10.759 --> 00:11:13.080 what does this all mean for you listening to 00:11:13.080 --> 00:11:16.960 this? We started with this huge high level strategy. 00:11:17.340 --> 00:11:19.559 AI will save the world. And we ended with a really 00:11:19.559 --> 00:11:22.659 powerful practical blueprint. Yeah, AI is this 00:11:22.659 --> 00:11:25.559 world changing revolution. But the real results, 00:11:25.820 --> 00:11:28.340 like cutting organizational friction by 50 percent 00:11:28.340 --> 00:11:31.340 in six months, that comes from disciplined execution. 00:11:31.440 --> 00:11:33.779 and the mundane work of process simplification. 00:11:34.019 --> 00:11:36.440 The two core takeaways are really clear. First, 00:11:36.580 --> 00:11:38.799 all that executive excitement about AI has to 00:11:38.799 --> 00:11:42.000 be backed by tangible programs, financial incentives, 00:11:42.299 --> 00:11:44.480 like that mastermind challenge. Right, to harness 00:11:44.480 --> 00:11:46.919 that bottom -up innovation. And second, these 00:11:46.919 --> 00:11:50.100 advanced AI tools are not magic. They need the 00:11:50.100 --> 00:11:52.799 traditional discipline of process review to have 00:11:52.799 --> 00:11:56.360 maximum impact. The AI only shines when the process 00:11:56.360 --> 00:11:59.559 underneath it is already efficient. So the CIOs 00:11:59.559 --> 00:12:01.879 we listened to basically established that professional 00:12:01.879 --> 00:12:04.039 growth, whether it's for an organization or for 00:12:04.039 --> 00:12:07.139 yourself, it's a continuous, never -ending journey 00:12:07.139 --> 00:12:09.659 toward a higher vista. Which leaves a final thought. 00:12:10.100 --> 00:12:12.279 What area of your own learning or your professional 00:12:12.279 --> 00:12:14.080 development are you currently treating like a 00:12:14.080 --> 00:12:16.240 destination, a mountain you can just conquer 00:12:16.240 --> 00:12:18.500 and be done with, when really it should be viewed 00:12:18.500 --> 00:12:20.919 as a forever thing? Something to think about 00:12:20.919 --> 00:12:22.320 as you get ready for your next climb.