Welcome Back - Featured Keynotes

Day 1 · Oct 13 Day 1 digest · More from this day

MC Alex Lieberman returns after the afternoon break to thank the sponsor tiers, argues the conference content is “surface area” for hallway conversations, shares Tenex’s enterprise AI diffusion ladder (90% LLM rollout, 60% trained users, 50% coding agents, 5% non-engineering agents), and introduces Frank Long of Coatue for “A Hedge Fund as a Function.”

Read the official edition ↗ A detailed reading edition with a searchable transcript.

TL;DR

• Lieberman resumes mid-thought (“For Fortune 500 businesses”) after a roughly 10-minute untranscribed gap, checks in on the day, and thanks the sponsors by tier: presenting sponsor Arize; platinum sponsors Oracle, Sonar, WorkOS, and Exa; then gold, silver, and supporting partners.
• He tells the room the talks themselves are just the starting point: the content, speakers, and depth are “surface area” for the conversations attendees have with each other.
• Recounting a backstage conversation with Swix, he shares Tenex’s enterprise AI diffusion ladder from roughly 100 enterprise clients: 90% have rolled out a general-purpose LLM to all employees, about 60% have people actually trained on the tools, around 50% use coding agents in engineering and product workflows, and only about 5% have agents working in non-engineering workflows.
• He runs the first stat as an interactive audience poll (“Someone scream out a number”) and gets guesses of 9%, 17%, and 90% before revealing the answer.
• He names the day’s recurring hallway themes: the diffusion gap, open source versus fine-tuned models, and how hard it is to hire great AI engineers at scale.
• He closes the block by introducing Frank Long, head of AI at Coatue (rendered “Kotu” in the transcript), for the talk “A Hedge Fund as a Function.”

Key lessons

1. Tenex’s diffusion ladder for enterprise AI, from about 100 enterprise conversations across financial services, healthcare, and industrials: 90% have adopted and rolled out a general-purpose LLM to all employees, but only about 60% have employees actually trained in using the technology, about 50% use coding agents in engineering and product workflows, and roughly 5% have agents working in non-engineering workflows.
2. Lieberman’s takeaway from the ladder: enterprises are “truly still day one” on the hardest milestone — agents in non-engineering workflows — which is the gap this community’s topics will fill over the next 12, 24, and 36 months.
3. His conference method: treat talks as surface area, not endpoints. Use each session as a prompt for conversations “at the edge” with other attendees, which is where the value of an in-person event lives.
4. Interactive stat delivery works on a live stage: he asks the room to scream out a percentage before revealing the 90% LLM-rollout figure, turning a data point into audience engagement.
5. The sponsor structure of AI Engineer NYC 2026 day one: presenting sponsor Arize; platinum sponsors Oracle, Sonar, WorkOS, and Exa; plus gold sponsors (“They’re all my favorite children”), silver sponsors, and supporting partners.
6. The hallway track’s dominant themes on day one, per Lieberman: the enterprise diffusion gap, open source versus fine-tuned models, and the difficulty of hiring great AI engineers at scale.

Watch this session

Alex Lieberman · 3:57:57 PM ET–4:02:23 PM ET · announcement · Watch on YouTube from 07:10:05

Notable quotes

“So before we hop into the next block, one thing that I just wanted to share, and I’m sure a lot of you are already feeling this already, is not only is the energy palpable across the room, in between all of the sessions, but what you realize when you come to AI Engineer is that the content is incredible.” [07:11:17 | 3:59:08 PM]

“We talk to hundreds of big enterprises at 10x across financial services, healthcare, industrials, et cetera.” [07:12:30 | 4:00:21 PM]

“And then the final one, where we are truly still day one, is agents actually working in non-engineering workflows in an organization. That is around 5% right now.” [07:13:33 | 4:01:25 PM]

“With that, I would love to introduce you to our next talk, which is with Frank Long, the head of AI at Kotu.” [07:14:20 | 4:02:11 PM]

Session narrative

This segment picks up the “Welcome Back - Featured Keynotes” block that Alex Lieberman opened briefly at 06:59:40 (session 040), after a roughly 10-minute untranscribed gap (session 041). He returns mid-sentence with “For Fortune 500 businesses,” finishes the thought he left hanging, and checks in with the room: most of the first session day is done, and he hopes the mix of programming and conversation has been fruitful. Note that the published agenda places this Welcome Back block at 4:25 PM; on the actual livestream it ran at 3:57:57 PM ET, so the stream ran ahead of the agenda.

The first order of business is the sponsor applause round, delivered by tier: presenting sponsor Arize gets the first shout-out, followed by platinum sponsors Oracle, Sonar, WorkOS, and Exa, then gold sponsors (he jokes he cannot read them all off because “They’re all my favorite children”), silver sponsors, and the supporting partners. A quick crowd aside follows, where he remarks on the energy of an attendee.

The core of the segment is reflection plus data. Lieberman argues that what makes AI Engineer work is that the curated talks act as “surface area” for attendee conversations — he has spent the day standing in the back of the room listening to speakers who are “truly living in Edge City,” then using those talks as openers for hallway discussions. He recounts one such conversation with Swix about the diffusion gap in enterprise AI, and turns it into an interactive moment: he asks the room to guess what percentage of about 100 enterprises have rolled out a general-purpose LLM to all employees, hears 9%, 17%, and 90% from the crowd, and confirms 90%. He then walks down the rest of the ladder — about 60% trained users, around 50% coding agents in engineering workflows, and about 5% for agents in non-engineering workflows — closing the loop on the day’s “Edge City” framing: the topics on this stage will play out across enterprises over the next 12, 24, and 36 months. He names the day’s other recurring conversation topics (open source versus fine-tuned models, hiring great AI engineers at scale) and urges attendees to keep using the content as a conversation starter.

He ends the block by introducing the next talk: Frank Long, head of AI at Coatue (the transcript spells it “Kotu”), speaking on “A hedge fund as a function” — which becomes session 043.

Speaker details

Alex Lieberman - Managing Partner, Tenex. Serial entrepreneur best known as co-founder of Morning Brew, the business media company he started as a college newsletter and grew to 2M+ daily readers before its ~$75M sale to Business Insider in 2020. He is now co-founder and Managing Partner of Tenex, an AI engineering and transformation firm that helps companies go from AI-absent to AI-native. He also founded Storyarb and Distro, hosts the Founder’s Journal podcast, and co-hosts The Crazy Ones. At AI Engineer NYC 2026 he served as MC, delivering the welcome, keynote intros, and closing remarks. In this segment he self-identifies through the line “We talk to hundreds of big enterprises at 10x across financial services, healthcare, industrials, et cetera.”

Links: LinkedIn | Twitter/X | Website

Full timestamped transcript

[07:10:05 | 3:57:57 PM] For Fortune 500 businesses.
[07:10:08 | 3:57:59 PM] And we are most of the way through the first session today, and hopefully it's been filled with incredible conversation, which I'm going to share a little bit about some of the conversations I've had offstage in a minute.
[07:10:22 | 3:58:13 PM] But hopefully the mix of the programming, the conversation you've had amongst everyone has been really fruitful today.
[07:10:29 | 3:58:20 PM] Now, before we hop into the final block of the day, let's just take a moment again to give a round of applause to our partners who make this entire event possible.
[07:10:39 | 3:58:31 PM] So let's start with our presenting sponsor, Arize.
[07:10:43 | 3:58:34 PM] Let's give a shout out to Arize.
[07:10:46 | 3:58:37 PM] Yeah, Arize.
[07:10:47 | 3:58:39 PM] Next up, we have our platinum sponsors, Oracle, Sonar, WorkOS, and Exa.
[07:10:53 | 3:58:44 PM] Let's give a shout out for them.
[07:10:55 | 3:58:47 PM] And then two more.
[07:10:56 | 3:58:48 PM] We have gold sponsors.
[07:10:58 | 3:58:49 PM] I love them all.
[07:10:58 | 3:58:50 PM] They're all my favorite children.
[07:11:00 | 3:58:51 PM] Can't read them all off right now, but thank you so much.
[07:11:04 | 3:58:55 PM] Our silver sponsors, and to all of our supporting partners who make this possible.
[07:11:10 | 3:59:01 PM] Thank everyone.
[07:11:11 | 3:59:03 PM] Yeah, I don't know if that was the first person who said they have the most energy, but that person has a lot of energy.
[07:11:17 | 3:59:08 PM] So before we hop into the next block, one thing that I just wanted to share, and I'm sure a lot of you are already feeling this already, is not only is the energy palpable across the room, in between all of the sessions, but what you realize when you come to AI Engineer is that the content is incredible.
[07:11:38 | 3:59:30 PM] The speakers that are curated are incredible.
[07:11:41 | 3:59:33 PM] The depth by which they talk about the work they're doing is incredible.
[07:11:45 | 3:59:36 PM] But this is also just surface area that acts as a starting point for conversation amongst everyone in the room.
[07:11:52 | 3:59:43 PM] And so for me, not only has it been incredible to stand in the back of the room at various points in the day and just listen to these incredible speakers who are truly living in Edge City, but then to use all of these conversations, all of these talks as surface area to have conversations with folks.
[07:12:10 | 4:00:02 PM] So a few conversations that come to mind that I've had throughout the day because of the content.
[07:12:16 | 4:00:07 PM] One was a conversation with Swix, one of the people behind this incredible event.
[07:12:22 | 4:00:14 PM] We were talking about Just the diffusion gap that exists in enterprise AI right now.
[07:12:28 | 4:00:20 PM] And just some fun stats for you all.
[07:12:30 | 4:00:21 PM] We talk to hundreds of big enterprises at 10x across financial services, healthcare, industrials, et cetera.
[07:12:38 | 4:00:29 PM] And the way I think about diffusion of AI across these organizations is almost in milestones of a race.
[07:12:45 | 4:00:37 PM] So the first big milestone at an enterprise is just adopting a general purpose LLM.
[07:12:52 | 4:00:43 PM] Out of curiosity, of the, let's call it 100 firms we talked to, what percentage of those firms do you think have adopted and rolled out an LLM to all of their employees?
[07:13:01 | 4:00:53 PM] Someone scream out a number.
[07:13:04 | 4:00:56 PM] Nine percent, I think I heard.
[07:13:07 | 4:00:59 PM] Seventeen percent.
[07:13:08 | 4:01:00 PM] So specific.
[07:13:10 | 4:01:01 PM] Ninety percent.
[07:13:11 | 4:01:03 PM] So 90% of the enterprises we work with have rolled out a general purpose LLM.
[07:13:17 | 4:01:08 PM] Under that is people actually getting trained in how to use the technology, how to actually use these surfaces in their work.
[07:13:23 | 4:01:14 PM] That's like 60%.
[07:13:25 | 4:01:16 PM] Next up is coding agents, agents that are actually used within engineering and product workflows.
[07:13:30 | 4:01:22 PM] That's around 50% of all the enterprises we've been in.
[07:13:33 | 4:01:25 PM] And then the final one, where we are truly still day one, is agents actually working in non-engineering workflows in an organization.
[07:13:43 | 4:01:34 PM] That is around 5% right now.
[07:13:45 | 4:01:37 PM] So going back to what I said earlier, this community really is edge city.
[07:13:50 | 4:01:41 PM] The topics that are being talked about, we are going to continue to see those play out over the next 12, 24, 36 months.
[07:13:58 | 4:01:50 PM] I've had a variety of conversations like this throughout the day: diffusion gap, open source versus fine-tuned models, how hard it is to hire great AI engineers at scale.
[07:14:09 | 4:02:01 PM] And so as we go about the rest of today, but also as you think about the conference in general, use the content as a starting point to have interesting conversations at the edge.
[07:14:20 | 4:02:11 PM] With that, I would love to introduce you to our next talk, which is with Frank Long, the head of AI at Kotu.
[07:14:28 | 4:02:19 PM] He's going to be talking about a hedge fund as a function.
[07:14:30 | 4:02:22 PM] Let's welcome Frank.