1 00:00:04,926 --> 00:00:16,946 Good morning and welcome back. We're excited to spend this final day together building on the conversations already underway and making room for a few more big conversation. 2 00:00:16,946 --> 00:00:24,346 Without further ado, please welcome today's emcee, founder, writer, and analyst at Turing 3 00:00:24,346 --> 00:00:26,146 Post, Ksenia Se. 4 00:00:35,306 --> 00:00:37,906 Good morning, good morning, good morning, everyone. 5 00:00:39,206 --> 00:00:43,106 Welcome to the third and the final day of AI Engineer New York. 6 00:00:43,306 --> 00:00:45,386 Who went to a cool event last night? 7 00:00:45,586 --> 00:00:46,426 What was the coolest? 8 00:00:48,386 --> 00:00:52,086 Okay, I went to karaoke with swyx, and it was amazing. 9 00:00:52,286 --> 00:00:53,206 100% recommend. 10 00:00:54,346 --> 00:00:56,686 So, yes, I'm Ksenia Se. 11 00:00:56,686 --> 00:01:00,086 I write Turing Post, and I'm your emcee for today. 12 00:01:00,446 --> 00:01:04,406 It's actually my second time emceeing AI engineer in New York, 13 00:01:04,586 --> 00:01:06,306 so that's how much I love it. 14 00:01:06,806 --> 00:01:10,266 And let's go to the program. 15 00:01:10,606 --> 00:01:14,286 Because yesterday, Aparna Dhinakaran from Arize 16 00:01:14,606 --> 00:01:17,506 asked, what does an AI engineer do? 17 00:01:17,866 --> 00:01:20,686 And today we have a session from Ryan Carson 18 00:01:20,686 --> 00:01:23,926 called Your Job is Now Agent Manager. 19 00:01:24,166 --> 00:01:28,046 So apparently your job description just changed overnight, 20 00:01:28,146 --> 00:01:30,726 and I hope that the compensation will keep up. 21 00:01:30,986 --> 00:01:34,466 But fortunately, quite a few of your employers 22 00:01:37,446 --> 00:01:38,966 are also our sponsors. 23 00:01:39,306 --> 00:01:44,146 So they've seen how quickly your responsibility 24 00:01:44,146 --> 00:01:46,746 changing. And let's give 25 00:01:46,746 --> 00:01:48,886 them a warm round of applause. 26 00:01:49,266 --> 00:01:50,526 Little encouragement 27 00:01:50,526 --> 00:01:52,306 goes a long way. 28 00:01:52,686 --> 00:01:53,986 Our most amazing 29 00:01:53,986 --> 00:01:56,626 presenting sponsor, Arize, please. 30 00:01:58,646 --> 00:02:00,566 Thank you, thank you, thank you. 31 00:02:01,186 --> 00:02:02,786 The great four, our platinum 32 00:02:02,786 --> 00:02:04,966 sponsors, Oracle, Sonar, 33 00:02:05,246 --> 00:02:06,466 WorkOS, and 34 00:02:06,466 --> 00:02:08,666 AXA, please, big round of 35 00:02:08,666 --> 00:02:09,086 applause. 36 00:02:11,086 --> 00:02:12,326 Scattering gold 37 00:02:12,326 --> 00:02:15,826 of gold sponsors all on this screen. 38 00:02:16,446 --> 00:02:19,326 Please, let's acknowledge everyone with our eyes 39 00:02:19,326 --> 00:02:20,926 and give them a big hand. 40 00:02:23,006 --> 00:02:25,186 And our silver sponsors, 41 00:02:25,746 --> 00:02:30,146 let's give them as much applause as possible 42 00:02:30,146 --> 00:02:33,126 because we don't raise enthusiasm by fear. 43 00:02:33,486 --> 00:02:34,766 So please, clap, clap, clap, clap, clap. 44 00:02:35,666 --> 00:02:37,626 And finally, to all of us, 45 00:02:37,726 --> 00:02:39,266 to all our supporting partners, 46 00:02:39,546 --> 00:02:40,826 thank you for making this possible. 47 00:02:40,826 --> 00:02:43,026 It would not be possible without anyone. 48 00:02:44,146 --> 00:02:46,886 Like everyone, you know, it takes a village. 49 00:02:47,146 --> 00:02:48,926 So thank you, thank you, thank you. 50 00:02:50,246 --> 00:02:51,526 I'm a little nervous. 51 00:02:51,526 --> 00:02:54,346 It's early, but thank you for bearing with me. 52 00:02:55,446 --> 00:02:59,126 If you've noticed, there is an interesting progression 53 00:02:59,126 --> 00:03:00,926 across all these three days. 54 00:03:01,186 --> 00:03:05,226 On Mondays, you could join super practical hands-on workshops. 55 00:03:05,566 --> 00:03:09,626 And you heard from big guys like Cerebras and OpenAI 56 00:03:09,626 --> 00:03:12,946 to small but super interesting startups like OpenPros. 57 00:03:13,306 --> 00:03:20,546 Yesterday, the conversation included how AI translates into operating margin. 58 00:03:21,066 --> 00:03:25,846 And today, we'll hear about the capabilities companies are building toward. 59 00:03:26,226 --> 00:03:32,206 The work agents can take on and what managing that might involve. 60 00:03:32,646 --> 00:03:36,786 Things will be happening here in Shannon Room. 61 00:03:36,786 --> 00:03:39,946 and it will be finance, compute, and infrastructure sessions. 62 00:03:40,226 --> 00:03:42,686 In Turing room, my favorite name, 63 00:03:43,026 --> 00:03:45,446 evaluations, benchmarking, retrieval, 64 00:03:45,846 --> 00:03:47,306 self-improving agents and skills. 65 00:03:47,686 --> 00:03:51,326 In Saturn room, you will hear about developer tools, 66 00:03:51,946 --> 00:03:56,406 enterprise data, software factories, 67 00:03:56,986 --> 00:03:59,066 agent permissions, and token economics. 68 00:03:59,486 --> 00:04:02,986 In McCarthy room, finally, will be about model costs, 69 00:04:03,386 --> 00:04:05,566 coding agent context, human verification, 70 00:04:05,566 --> 00:04:09,606 and banking deployment and financial agent evidence. 71 00:04:10,506 --> 00:04:13,946 And at Turing Post, we spend a lot of time 72 00:04:13,946 --> 00:04:17,806 following all these directions 73 00:04:17,806 --> 00:04:19,466 and research in these directions. 74 00:04:20,006 --> 00:04:25,446 But we also have a series called the Org Age of AI, 75 00:04:25,446 --> 00:04:28,966 where we look at what happens to a company 76 00:04:28,966 --> 00:04:31,266 as it begins to work with this technology. 77 00:04:31,766 --> 00:04:33,826 How does the work change, right? 78 00:04:33,826 --> 00:04:37,126 What becomes possible that wasn't practical before? 79 00:04:37,466 --> 00:04:41,826 And while superintelligence feels so powerful, 80 00:04:42,266 --> 00:04:44,886 why is the ROI still missing? 81 00:04:45,326 --> 00:04:49,926 And that's part of what makes this gathering so interesting to me. 82 00:04:50,446 --> 00:04:53,046 You have people developing the technology 83 00:04:53,046 --> 00:04:57,106 in the same room as people who understand in considerable detail 84 00:04:57,106 --> 00:04:58,426 the work it needs to do. 85 00:04:58,606 --> 00:05:02,206 And there's a lot of those two groups can learn from each other. 86 00:05:02,206 --> 00:05:03,766 That's fascinating to me. 87 00:05:04,026 --> 00:05:05,026 And our 88 00:05:05,226 --> 00:05:11,126 first speakers work at the intersection of those two worlds, at Ramp. 89 00:05:11,426 --> 00:05:18,866 So please help me welcome Alexander Shevchenko with Chasing Capabilities While Waiting for the Frontier. Thank you.