WEBVTT 1 00:00:00.250 --> 00:00:00.875 Hello, Manoj. 2 00:00:00.875 --> 00:00:02.419 Lovely to see you in Davos. 3 00:00:02.836 --> 00:00:04.754 Likewise, Vaishali, pleasure to see you. 4 00:00:05.171 --> 00:00:05.922 Fantastic. 5 00:00:06.423 --> 00:00:07.716 It's an honor to have you today. 6 00:00:07.716 --> 00:00:11.177 I'm going to ask you a couple of questions, and of course the 7 00:00:11.177 --> 00:00:12.387 topic is AI and GenAI. 8 00:00:12.554 --> 00:00:13.722 I wouldn't have guessed. 9 00:00:15.223 --> 00:00:18.226 So first, how has AI shaped Juniper's business? 10 00:00:19.519 --> 00:00:22.230 We would like to think of ourselves, our company, as a 11 00:00:22.230 --> 00:00:24.315 company who is the keeper of the Internet. 12 00:00:24.315 --> 00:00:25.734 We build routers and switches. 13 00:00:26.276 --> 00:00:28.903 We've been part of the Internet revolution, you know, which 14 00:00:28.903 --> 00:00:31.197 happened about 25 years ago, Internet and smartphone 15 00:00:31.197 --> 00:00:31.698 revolution. 16 00:00:31.948 --> 00:00:34.868 We powered the networks, which made all of this possible. 17 00:00:35.326 --> 00:00:38.705 And now I believe AI is that next inflection point, in some 18 00:00:38.705 --> 00:00:41.374 ways even bigger than the Internet revolution. 19 00:00:41.958 --> 00:00:46.087 So for Juniper, we've been investing in AI for the last six 20 00:00:46.087 --> 00:00:47.130 to seven years. 21 00:00:47.422 --> 00:00:51.176 We had invested in AI Ops specifically with one thing in 22 00:00:51.176 --> 00:00:51.468 mind: 23 00:00:51.926 --> 00:00:53.261 The networks are too complex. 24 00:00:53.595 --> 00:00:56.181 You know, it doesn't need to be, you know, so you don't need a 25 00:00:56.181 --> 00:00:57.182 PhD to manage a network. 26 00:00:57.182 --> 00:00:59.184 You don't need to spend like years to manage a network. 27 00:00:59.434 --> 00:01:00.351 It's got to be simple. 28 00:01:00.643 --> 00:01:03.605 So about six to seven years ago, we acquired a company called 29 00:01:03.605 --> 00:01:06.399 Mist and then we 'Mist-ified' the entire Juniper portfolio 30 00:01:06.399 --> 00:01:07.233 with that company. 31 00:01:07.609 --> 00:01:12.530 And what they did was an amazing thing, a cloud-native, fully 32 00:01:12.530 --> 00:01:17.702 microservices-based AI cloud and an AI engine together which can 33 00:01:17.702 --> 00:01:20.580 look at all aspects of the network. 34 00:01:20.705 --> 00:01:22.165 You know, Day Zero. 35 00:01:22.332 --> 00:01:26.044 You just scan the barcode, QR code, and things are up and 36 00:01:26.044 --> 00:01:29.422 running, and then the whole network self-discovers. 37 00:01:29.589 --> 00:01:32.717 And then the best part is when it operates, it's a self-healing 38 00:01:32.717 --> 00:01:33.134 network. 39 00:01:33.259 --> 00:01:35.637 I was just thinking self-healing actually, yes. 40 00:01:36.179 --> 00:01:39.516 And the self-healing comes because we are able to collect 41 00:01:39.516 --> 00:01:43.103 about 15 to 20 data points on a per-second-basis from each of 42 00:01:43.103 --> 00:01:46.731 the Wi-Fi equipment, the LAN equipment, or the WAN equipment, 43 00:01:46.731 --> 00:01:50.235 and put them all together and the AI engine starts learning. 44 00:01:50.235 --> 00:01:52.904 Over a period of six to seven years now, we've become so 45 00:01:52.904 --> 00:01:55.615 proficient that 90% of the problems can be solved before 46 00:01:55.615 --> 00:01:56.157 they happen. 47 00:01:56.449 --> 00:01:59.202 ServiceNow is one of our largest customers, 90% of the trouble 48 00:01:59.202 --> 00:02:01.913 tickets have been eliminated and that's one of the sea change 49 00:02:01.913 --> 00:02:03.832 which is driving this evolution of Juniper. 50 00:02:03.832 --> 00:02:06.209 So Juniper has been in the midst of AI and now GenAI 51 00:02:06.209 --> 00:02:09.379 puts that into, you know, over speed. 52 00:02:09.546 --> 00:02:10.255 Over speed, exactly. 53 00:02:11.214 --> 00:02:11.464 Wow. 54 00:02:11.464 --> 00:02:15.343 I had actually no idea that, you know, you were so much along in 55 00:02:15.343 --> 00:02:18.972 the journey and actually what you're talking about is simply 56 00:02:18.972 --> 00:02:20.056 fantastic to hear. 57 00:02:20.056 --> 00:02:22.851 So I'm going to ask you, if you've done so much. 58 00:02:23.101 --> 00:02:26.104 It definitely has had an impact on your workforce. 59 00:02:26.104 --> 00:02:29.524 Can you talk a little bit about that or also how are you using 60 00:02:29.524 --> 00:02:31.276 technology to impact operations? 61 00:02:32.026 --> 00:02:34.946 Love to, I think I'll start with customers always, you know, 62 00:02:34.946 --> 00:02:36.239 start with customers first. 63 00:02:36.281 --> 00:02:38.950 If you look at our customer base, you know it starts over 64 00:02:38.950 --> 00:02:41.661 with, you know, even something as simple as like you cannot 65 00:02:41.661 --> 00:02:42.745 connect to the network. 66 00:02:43.037 --> 00:02:44.497 You immediately blame the Wi-Fi. 67 00:02:44.664 --> 00:02:46.791 Wi-Fi sucks. 68 00:02:47.000 --> 00:02:48.209 That's what people say, right? 69 00:02:48.418 --> 00:02:49.961 But it may have nothing to do with Wi-Fi. 70 00:02:49.961 --> 00:02:51.671 It may have something to do with the faulty cable. 71 00:02:51.921 --> 00:02:53.673 It may have something to do with the switch, it may have 72 00:02:53.673 --> 00:02:55.258 something to do with the Internet connection to the 73 00:02:55.258 --> 00:02:55.758 service provider. 74 00:02:56.092 --> 00:02:59.512 So we realized, OK, what we need to do is make things easy for 75 00:02:59.512 --> 00:03:02.891 the customers, easy to operate, and for the end user like you, 76 00:03:02.891 --> 00:03:06.019 Vaishali, or your Zoom, or your team's call, it should be 77 00:03:06.019 --> 00:03:06.519 flawless. 78 00:03:07.020 --> 00:03:08.521 It doesn't matter what the problem is, we should 79 00:03:08.521 --> 00:03:09.022 self-correct it. 80 00:03:09.022 --> 00:03:12.442 So with that vision in mind, our customers found it like, wow, 81 00:03:12.442 --> 00:03:15.653 you know, things just work and if there's something going 82 00:03:15.653 --> 00:03:17.739 wrong, you're going to fix it itself. 83 00:03:17.989 --> 00:03:20.909 Which means that I have time now to do important things. 84 00:03:21.201 --> 00:03:23.411 Previously I was doing the mundane work, now I can do 85 00:03:23.411 --> 00:03:24.078 important things. 86 00:03:24.287 --> 00:03:26.956 Similarly for employees, our engineers, they're spending less 87 00:03:26.956 --> 00:03:29.000 time debugging stuff because they have all the 88 00:03:29.000 --> 00:03:31.002 self-correction and auto-correction happening. 89 00:03:31.419 --> 00:03:33.880 And even in cases where it's something new that's happening, 90 00:03:33.880 --> 00:03:36.507 they can figure it out and train it, and the next time they don't 91 00:03:36.507 --> 00:03:37.759 need to worry about it anymore. 92 00:03:38.259 --> 00:03:41.221 And for product managers, it's amazing because you know when 93 00:03:41.221 --> 00:03:44.390 customers are using a particular feature, when customers are not 94 00:03:44.390 --> 00:03:47.560 using it, what's useful, what's making lots of money, what's not 95 00:03:47.560 --> 00:03:48.603 making lots of money. 96 00:03:48.603 --> 00:03:52.565 So the cycle is reduced to such a short time, you know what to 97 00:03:52.565 --> 00:03:53.691 build much faster. 98 00:03:53.691 --> 00:03:55.818 You don't need to wait for customer feedback and surveys 99 00:03:55.818 --> 00:03:56.527 and all that stuff. 100 00:03:56.819 --> 00:04:00.156 So it speeds up innovation and development and last but not the 101 00:04:00.156 --> 00:04:03.159 least, you know, eat your own dog food or bring your own 102 00:04:03.159 --> 00:04:04.035 champagne, right? 103 00:04:04.035 --> 00:04:05.286 And it makes it super easy. 104 00:04:05.286 --> 00:04:07.830 Our entire network is running on you know, our stuff. 105 00:04:08.289 --> 00:04:11.334 And now with GenAI coming, there's a lot more new 106 00:04:11.334 --> 00:04:11.960 frontiers. 107 00:04:12.335 --> 00:04:15.088 You know, we can do Copilots, you know, to speed up 108 00:04:15.088 --> 00:04:17.715 development, testing, lots of things to be done. 109 00:04:17.715 --> 00:04:20.593 So we're just embracing AI in a full way. 110 00:04:20.760 --> 00:04:24.180 And from legal documents, to running networks, there's a lot 111 00:04:24.180 --> 00:04:25.306 of stuff to be done. 112 00:04:26.182 --> 00:04:27.267 Wow, that's impressive. 113 00:04:27.267 --> 00:04:29.978 And to bring the, you know, employees along this journey, 114 00:04:29.978 --> 00:04:32.897 because I'm sure it would have required upskilling, retraining 115 00:04:32.897 --> 00:04:33.273 as well. 116 00:04:33.273 --> 00:04:34.857 I'm sure you did all of that too. 117 00:04:34.857 --> 00:04:35.400 Correct? 118 00:04:35.483 --> 00:04:36.818 We're still in the process of that. 119 00:04:37.026 --> 00:04:40.071 You know, I think, you know, this is new to everyone. 120 00:04:40.071 --> 00:04:42.949 Like one of the things which you have a dearth of is, you know, 121 00:04:42.949 --> 00:04:45.660 AI analysts and data scientists and whatnot, but everybody's 122 00:04:45.660 --> 00:04:46.119 trainable. 123 00:04:46.202 --> 00:04:48.705 I think one of the good things about in smart people is that 124 00:04:48.705 --> 00:04:51.165 they can move from discipline one to discipline two if you 125 00:04:51.165 --> 00:04:52.709 provide the right amount of training. 126 00:04:52.709 --> 00:04:54.294 So investment is important. 127 00:04:54.544 --> 00:04:55.878 So you're not going to be left behind. 128 00:04:56.129 --> 00:04:58.256 You don't have need to have any fear of missing out. 129 00:04:58.464 --> 00:05:01.259 You just need to have the right attitude and the aptitude. 130 00:05:01.509 --> 00:05:04.804 I like it, the right attitude and the right aptitude, I like 131 00:05:04.804 --> 00:05:05.388 that a lot. 132 00:05:05.763 --> 00:05:09.434 So maybe in conclusion, any learnings for you know other 133 00:05:09.434 --> 00:05:13.313 companies to take away, you know based on in the fact as you 134 00:05:13.313 --> 00:05:16.983 said, you know, you're doing everything from what I call 135 00:05:16.983 --> 00:05:20.820 productivity in the back office, to really your own product 136 00:05:20.820 --> 00:05:21.904 suite, to beyond. 137 00:05:22.488 --> 00:05:25.199 So any learnings, I mean if people are just starting this 138 00:05:25.199 --> 00:05:26.826 journey, what would you recommend? 139 00:05:27.243 --> 00:05:30.079 I think at a minimum, people should automate 140 00:05:30.079 --> 00:05:34.083 processes—whether it's you know, code to cache, things from the 141 00:05:34.083 --> 00:05:37.420 back office to the front end, chat bots for customer 142 00:05:37.420 --> 00:05:39.547 service—automation is number one. 143 00:05:40.131 --> 00:05:43.801 Once you do automation, I think using AI insights to proactively 144 00:05:43.801 --> 00:05:47.347 solve problems, reduce all this workload, and actually improve 145 00:05:47.347 --> 00:05:48.097 productivity. 146 00:05:48.306 --> 00:05:50.808 And also, if I may say one thing, it's actually good for 147 00:05:50.808 --> 00:05:53.561 sustainability too, because if you solve these problems, you're 148 00:05:53.561 --> 00:05:54.645 avoiding the truck roles. 149 00:05:54.979 --> 00:05:56.356 So it's good for a better planet, too. 150 00:05:56.606 --> 00:06:01.194 So start small, but start with automation, engage in AI, GenAI, 151 00:06:01.235 --> 00:06:03.237 and the world is wide open. 152 00:06:03.529 --> 00:06:06.199 So start small but have a bigger ambition. 153 00:06:06.199 --> 00:06:06.949 So absolutely. 154 00:06:06.991 --> 00:06:08.409 So thank you Manoj. 155 00:06:08.493 --> 00:06:12.455 That was very insightful and really good to have you here in 156 00:06:12.455 --> 00:06:12.830 Davos. 157 00:06:12.830 --> 00:06:14.248 And enjoy the snow. 158 00:06:14.624 --> 00:06:15.333 Thank you, Vaishali, 159 00:06:15.333 --> 00:06:16.167 it's a pleasure to be here. 160 00:06:16.334 --> 00:06:16.793 Thank you.