Eno Thereska on Agentic Security and DevRel

00:00 Hey, Tim Berglund here on

00:01 my own occasional podcast.

00:04 My guest today on this

00:04 first episode is Eno Tereska.

00:07 He and I briefly overlapped at Confluent,

00:09 and then he went on to AWS, Alcion, Veeam,

00:13 and basically being an elite

00:15 engineer in all those places,

00:17 including Confluent.

00:19 Now he's a first time founder running a

00:21 company called Trent,

00:23 a startup focused on

00:23 security for AI agents.

00:27 We get into some of the tensions

00:28 that are really at the

00:29 heart of agentic development

00:30 right now in these early

00:31 days, which is that, sure, fine,

00:33 we're faster than ever.

00:34 I think everybody agrees with that.

00:36 But things like security

00:37 practices haven't caught up,

00:40 and there's good reason to

00:41 think that they might never.

00:42 That's a gap that Eno's

00:43 startup is, of course,

00:44 trying to fill.

00:46 We also talk about DevRel

00:47 in early stage startups,

00:48 when to make that first hire,

00:50 the importance of trust

00:51 in DevRel in the age of AI.

00:55 Questions like, is a developer advocate

00:56 the same as a field CTO?

00:59 All this and more, let's jump in.

01:02 Hey, you how's it going?

01:04 Good, good to see you.

01:05 It's been a while.

01:07 It has it has.

01:09 were uh talking recently and uh I just I realized at the time it would be better if we were recording the conversation.

01:19 It was just like I don't know.

01:21 It was a good conversation.

01:22 I thought we need to revisit this.

01:24 So uh let's do it.

01:27 But let's start off with uh for anybody doesn't know your background doesn't know what you're doing right now.

01:34 What are you doing right now.

01:36 How'd you get there?

01:37 Started a startup um in the agentic space, of course, um working on security.

01:45 We want to make those agents safe and secure.

01:49 A lot of my background, I've been an IC up to this point.

01:52 So this is a first time founder.

01:54 Previously, I worked with you, some at Confluent.

01:58 It's lucky to overlap for a couple of months.

02:00 Yeah, that was really great.

02:01 um Previously, I was an engineer at Veeam, at Alcion.

02:08 Spend some time in AWS, Confluent of course.

02:11 Always an IC

02:13 Coded a lot, still do.

02:16 Yeah, now everything has changed.

02:19 Probably for the second time.

02:22 I was there in the on-prem to cloud and now cloud to agentic.

02:26 Yeah, it's good to see more than one big shift.

02:30 I really think this is the biggest, we're around the same age, but in our professional lifetimes

02:36 I think this is the biggest shift and it's happened within the last five or six months

02:42 So we're still just kind of getting our feet under us with this changing landscape.

02:46 Yeah, yeah, no, I agree.

02:49 um, listen, I dropped everything for this.

02:51 I uh, I said everything is changing.

02:54 However, I did see some glimpses of this, uh, back when, uh, was working on my PhD at Carnegie Mellon, we built.

03:02 We built the first autonomous data center back then.

03:05 So this was, this was a long time ago.

03:07 I'm not going to mention the year around 2007.

03:10 I was gonna say, it's impolite to...

03:12 Yeah.

03:12 So back then we were using, I remember we took a

03:17 we took a cool...I'm a systems person, like an engineer, yeah, like I build stuff, but we were very excited.

03:24 We took this machine learning course from, think it was Tom Mitchell back then.

03:28 He's written that machine learning book and we were like, Hey, let's just use this.

03:32 Let's apply this.

03:33 But it wasn't LLMs.

03:35 was back then it was decision trees and neural networks and things like queuing theory.

03:39 And now it's...

03:41 LLMs and reinforcement learning.

03:44 But a lot of things are also similar.

03:46 I mean, back then we had mispredictions and errors and false positives, and now we have hallucinations.

03:52 um I had an early glimpse of this.

03:55 just, many years ago, I just didn't, at the same time, things have still moved really fast this last 12 months, really.

04:05 I am surprised.

04:06 Yeah.

04:08 uh Tell us a little bit about what Trent.ai is company, right?

04:10 I think it kind of frames the discussion about developer adoption and communities and things like that.

04:19 So, agentic security, but uh dig into that a little bit for me and let's talk about how

04:27 where developer adoption fits into that.

04:29 I know you got questions for me too, so.

04:32 Yeah, let's also keep this simple because "agentic" is a loaded term.

04:36 So here's how this started in pretty much every place I've been as a developer.

04:42 We were always disconnected from

04:45 We were always disconnected from any security teams.

04:46 So I think AWS was probably the best place that had some solidified security culture.

04:53 And there were shipping services.

04:54 We tried to do them really fast and ship on time.

04:59 And right at the last minute, this other team would show up, this security team

05:03 and they would say, like, hold on, you can't do any of this, Yeah.

05:10 Have you looked at this checklist?

05:12 And there was this big checklist of things, and we were like, okay, we just want to ship some stuff out there.

05:18 And they were very disconnected from the developers.

05:21 The developers were disconnected from them.

05:24 And, everybody tried really hard, but then it became this...

05:28 On one hand, we wanted a good security culture.

05:31 We wanted to, you know, we always said security first.

05:34 Yeah.

05:36 On the other hand, security always felt like

05:39 a bit of an afterthought.

05:41 And then in startups I've been to, um they just, don't even have a security team.

05:46 Yeah.

05:46 So people are shipping software really fast.

05:49 And I did some, I did some research back in the day and

05:54 It turns out 70% of security problems happen before you even write your first line of code.

05:59 So then, yeah, so it's like, if you get the design wrong, you get the architecture wrong, right?

06:05 And um how do we think about that?

06:08 What does that mean?

06:09 So we have interesting tools nowadays like Claude and, you know, we've heard of Mythos.

06:15 But what does it mean when you get security wrong before you even started writing the code?

06:21 So yeah, that got us thinking.

06:24 Fast forward to today, it's not about code.

06:28 It's more than code.

06:29 It's code, it's it's MCP servers, it's agents.

06:35 The definition of agents is in flux, things that are produced with Lang chain or open claw.

06:41 And it's quite interesting and quite complex.

06:44 But again, like going back to that culture of security, the question in our mind is, we're certainly shipping things really fast, but...

06:54 what does it mean to have them also secure and safe once they're deployed in production?

06:58 So this is what started the trend and um trying to help developers.

06:59 So this is what started the trend and um trying to...

07:04 ...help developers.

07:06 But it's this interesting question I'd love to explore with you, is on one hand,developers really shouldn't care about security.

07:12 They should just care about what they are excited about.

07:16 On the other hand, we have this tension that we do want

07:20 a culture where security does come first.

07:22 So we can explore a little bit the tension, especially nowadays where we're using so many tools

07:27 And in resolving the tension, we're kind of right at the heart of it, this trend.

07:32 uh I like it, I like it.

07:33 I guess on the one hand you don't want to have your solution be, well, let's just make everybody be good at security.

07:41 No, no, no, you don't want that. No one wants that.

07:44 well, mean, it would be nice, it's just that that's not available to you, you know.

07:47 ah It's more like what are the developers that you have and the processes they're using and, you know, how can you...

07:57 Find the levers that you've got then to make their systems more secure.

08:00 that sounds like, that sounds like, the second one sounds like.

08:03 Well, it's interesting.

08:05 can take an extreme there.

08:06 You talk with developers.

08:08 I was one of them.

08:09 And I mean, they don't get promoted based on security, right?

08:12 They get promoted based on features.

08:13 What did you launch?

08:15 How many agents did you launch?

08:18 And can we also, I mean, there's this challenge.

08:22 Can we make security invisible?

08:24 just, Trend, for example, produces an agent.

08:27 Maybe it just fixes the security for you.

08:30 But the problem is that if you go too far there, if you fix too much,

08:34 then what does it mean for the developers also learning about security

08:38 kind of appreciating those steps that are really important in the development of someone in their

08:45 software engineering career as well.

08:47 Yeah, so it's an interesting tension.

08:52 today, many people are skipping some of the intermediate steps.

08:55 They go from idea to final product.

08:58 Maybe they're using it, using lovable or bold to produce that.

09:02 Maybe they're vibe coding it.

09:03 So what do we do about this part in the middle is quite interesting.

09:09 And it needs to be done tastefully as well because um again, people have a job and they also want to go home after that.

09:16 Yeah.

09:16 And they don't think about security like all the time.

09:20 Right.

09:22 And as you say, they are more measured on, we'll just say features.

09:28 And in the current context, and I don't know how long this is going to last, but

09:33 many developers are, we'll just say, preoccupied with "What's my future?"

09:38 What does it even mean to be a person who writes software now?

09:41 uh And speaking for myself, last night, uh

09:48 adding some features to my personal website, which I'm building with Claude Code, having a great time.

09:54 um And it's not merely a static site.

09:56 There's a fair amount of actual code there.

09:59 And I was thinking, hey, comments.

10:01 I need people to give the ability for people to comment.

10:04 Let me write my own, which nobody would say.

10:07 It's not a chance you'd do a thing like that.

10:10 Now I'm going to.

10:12 I'm in the middle of it.

10:13 Got an OpenSpec proposal, looks good, about to start moving forward with it.

10:17 Have I thought about security?

10:19 ah No, I'm putting a tremendous amount of trust in Claude to get that right.

10:24 It's a fairly simple thing, it's next.js and the stack is fairly trustworthy, but

10:31 I'm being the normal developer who kind of doesn't think that hard about that

10:36 and needs that help

10:40 And it's, it's this interesting thing.

10:42 This it's not a paradox, but you have to hold two things simultaneously in mind

10:47 which is one, everything about security, everything we've learned over this last 50 years is now

10:55 encoded in, in an LLM.

10:58 So, um, low level code vulnerabilities, Anthropic is going to do it for you.

11:03 It's going to do a great job.

11:05 So you don't have to worry about them at the same time.

11:09 You know, it's opened up a whole bunch of new...

11:14 tech surface areas.

11:17 Just you have these agents that are doing a lot.

11:19 Maybe the code is, is correct, but all of a sudden they're touching parts of the system.

11:24 They're just not really supposed to touch.

11:26 So it's this interesting paradox where a lot of problems, a lot of the old problems in cybersecurity are now being addressed.

11:32 Um, and they're now fixed or fixable with relatively low cost and a whole bunch of new other problems are being opened.

11:44 And that's where there is the jump from code to

11:52 well, this code and this agent that you're deploying, it's really running and it's running with other agents.

11:58 And who really knows what it's doing?

12:00 I mean, you have a prompt to it and you give it your files, you give it your code to fix, right?

12:07 And where does that all go?

12:08 There's this...

12:11 You know, back in the day we had this, we had data and we had compute

12:15 I guess in Kafka, had like the Kafka topics and we had the records that we produced there.

12:20 And now once those things enter, once you do, once you develop your application and

12:26 once you have a prompt there that sucks the data in, all of a sudden the compute and the data

12:29 All of a sudden the compute and the data

12:30 have been mixed together.

12:32 The data is now part of the model and it's kind of embedded in there.

12:36 Yeah.

12:36 And teasing that the part is, it's actually quite hard.

12:40 So there's been this mashing of data and code, which I also find quite interesting.

12:46 potentially difficult in terms of solving problems like security.

12:50 So, where, uh maybe it's obvious, but how do you think about DevRel and security and Trent.ai and

13:03 talk to me about the, you know, where you think developers fit in into the

13:08 adoption of a thing like this in the place that we are right now.

13:10 It is very strange environment.

13:13 It is.

13:13 And here I also have a question for you because I work in two worlds and Trent works in two worlds.

13:19 m Maybe three worlds.

13:22 One of them is the community.

13:24 So we have a big play in the community.

13:28 So we have Trent for OpenClaw.

13:30 And there you have developers building these agents and deploying them.

13:33 And there we can talk about DevRel and getting people to understand the importance and the...

13:41 the joy and the discovery of building these agents and securing them.

13:47 And, and I can think of that as DevRel.

13:50 But then we also have an enterprise angle and there I'm kind of curious, what do you think

13:57 this might sound like a silly question, but what do you think about the

14:00 intersection of DevRel and something like a field CTO that we have more in the enterprise?

14:05 Because there, um, it's developers in the enterprise as well, but

14:10 They think more than just the code.

14:12 think more than just the tools.

14:14 They just think differently.

14:17 Maybe this is the first time I've asked anyone that question, right?

14:19 Like, do you say DevRel more like in the space where developers are directly working

14:25 with some of the concepts you've provided and then in the enterprise, what do you do?

14:31 Does it translate to a field CTO?

14:33 Is it like even a bad question?

14:35 Does the question even make sense?

14:36 No, it makes a lot of sense.

14:38 A developer advocate and a field CTO have a lot of similarities.

14:42 A field CTO is no, it's not stupid at all.

14:48 It's close to home.

14:51 So a field CTO is like a supercharged SE in the terms we'd use.

14:55 So sales engineers or systems engineers, solutions engineer, whatever, know, sales engineer, technical person.

15:01 And just for listeners, if you don't know the term,

15:03 technical person who's involved in the selling process glued to a salesperson to answer the technical questions of the selling process.

15:12 And a field CTO is like a real good version of that.

15:14 That's the elite.

15:16 And typically in an enterprise software company, a field CTO is in the go-to-market organization and maybe even in the sales organization and they're involved in selling.

15:29 And typically a developer advocate is not, you don't put like

15:33 quota numbers or anything like that on a dev rel team.

15:35 uh But the skill set is very similar.

15:40 like, at times I've, when you're recruiting developer advocates, say you find

15:44 somebody who's got technical skills or communication skills that look good, but they're not

15:49 currently a developer advocate, they might ask, where does my career go after this?

15:54 Is this a dead end?

15:55 Am I just this, you know, fluffy conference speaker for the rest of my life?

16:00 It's not.

16:01 what a developer advocate is, but people think that.

16:02 uh People worry that when they're coming in.

16:07 And I always say being an SE is adjacent.

16:10 You don't like this, just go be an SE.

16:11 You might make more money.

16:12 You know, you'll travel more.

16:13 It'll be more stressful, but pay is great.

16:16 And field CTO is a lot like that.

16:18 So if you're a person who understands the technology deeply and is good

16:24 with people, a good communicator, can write ideas together, tell a story.

16:30 the two sides of the same coin.

16:31 Not everybody wants to do both jobs, but they're very similar.

16:35 And correct me if I'm wrong, but one of the things that I'm seeing, I'd love to hear

16:39 what you're seeing there is that I'm seeing a lot of folks in the enterprise now with AI,

16:47 people that didn't necessarily code before, people like product managers.

16:52 um They are very open and curious to trying some of these new tools and to just playing around with it.

16:59 So we have people saying, Hey, I'm, I'm deploying the equivalent of open cloud.

17:04 There are different things in the enterprise and um all of a sudden I had, you know

17:10 I had these teams that were doing all the development, but all of a sudden as a product person,

17:13 I just want to put this thing together and do various things with it.

17:16 Like, you know, query all my data.

17:18 Right.

17:19 and so, so the folks that were talking before with field CDOs are now doing a lot of development and they're going through the journey of discovering.

17:31 Well, um you know, what are some of the tools I have?

17:34 What are some of the options I have?

17:35 Discovering also some of the joy of, hey, I'll try something small and then make it bigger.

17:41 Yeah.

17:43 And I'm finding that really fascinating because then...

17:45 um

17:49 You know, there are similarities to, to in that world in the enterprise with some of the

17:54 work that I'm seeing from what I call traditional developers more like in the OpenClaw community.

17:59 So that's that's been, that's been quite fascinating to watch that the software engineering.

18:04 it's a cliche to say it's been democratized, but it has led to this

18:11 this somewhat hybrid role that I'm seeing across the answer, like a hybrid, like DevRel

18:15 field CTO and product person in there.

18:19 So yeah, that's some of stuff that I'm seeing.

18:21 I was, you know, a year and a half ago that uh I started seeing our demos get a lot better

18:28 because, the typical talk a developer advocate is going to build, you know, you don't have

18:35 time to invest in a big fancy front end or like maybe typical for Kafka things.

18:41 There'll be some Grafana and you'll see some graphs or something, but you're just kind of connecting pipes there.

18:46 You're not going to build a custom front end or some cool

18:49 custom data generator that's part of the story you're telling.

18:52 You just don't have time for that.

18:53 Nobody's going to spend a month on a demo for a talk they're going to give five, 10 times.

18:59 Well, now we can move a lot faster on that stuff and have more fun there.

19:06 I built my first front end when we started the startup.

19:10 we didn't have someone in the front end.

19:12 So, so using V0, it's really good.

19:16 I, um, yeah, I can be dangerous now in that, in that, uh, it usually was more of a backend person, but, I can pretend to understand front end as well.

19:28 mean, um, the interesting thing is now, and I love your opinion there, there is this.

19:34 There is this gap between doing something really quickly and then deeply understanding it.

19:39 I don't think I understand front-end or the concepts there, but it has been easy using Vizier and other tools there.

19:45 I mean, we have a whole bunch of tools.

19:47 It's not just V0.

19:48 I started with it.

19:50 And one of the, one of the interesting things as we're building the startups, we're hiring various folks with various expertise is this.

20:02 almost like jumping of this middle layer where, you you start discovering something

20:06 then you learn a little bit more deeply, and then you get to that final product that you see

20:10 and you love.

20:12 It almost felt like I was cheating there a bit.

20:13 Like I started and then I got like almost like an end product, but I didn't really understand anything in the middle.

20:21 Or maybe I did, but not as deeply as if I had like handcrafted the code, right?

20:29 What's your thinking there in terms of the...

20:33 mean, are you seeing AI as kind of helping us?

20:38 That's a broad question.

20:40 Tell me a bit more about this like intermediate joy that people have in their discovery process.

20:44 Like have we like short-circuited that part or not?

20:49 I in terms of the joy, I don't think so at all.

20:53 I think that's that's there's it's very real.

20:56 And there's there's kind of two emotions uh that I hear come up a lot

21:03 when I listen to developers talk about agenetic development.

21:06 One of them is joy.

21:07 I can't believe I'm having much more fun than I've ever had.

21:11 I can't believe the kind of stuff I can build.

21:13 People say that.

21:14 That's me right now.

21:17 And then there's exhaustion.

21:19 I've got five agents running.

21:21 I can't keep track of it.

21:22 It's so addictive because I just get I think I just need to get another little move it forward a little bit.

21:27 I can't go to bed.

21:28 You know, so there's these these kind of two competing reactions.

21:32 Very interesting.

21:34 The second one, clearly, we're going to have to manage

21:37 we're going to build some some tools around learning when to stop.

21:44 I don't think that...

21:46 ...that joy is...

21:49 cheap.

21:50 There's nothing short-circuity about that.

21:52 think the making the thing that you want to make has always been what we're after.

21:59 And it just took a real long time and a lot more suffering to get there.

22:03 And now when you get to make something that much faster, you're still the...

22:10 And I'm speaking for myself and I think many software engineers—

22:16 It's been 15 years since I've earned my living as a software engineer.

22:19 OK, I'm it's this is the part of a developer that never goes away.

22:25 You still see yourself as a developer, even though it's been a long time.

22:27 But I think most of us, that's what we're in the game for, right, because we want to make things.

22:33 Some people are in the game because they really like crafting code.

22:38 And I think those people are sad right now.

22:40 And that's a that's a true grief, because that's not going to be a thing that

22:46 we get to do as much.

22:47 ah But I think most of us just like to make stuff and we get to make stuff faster.

22:53 Now the short circuiting or the short cutting

22:57 I think we are in a process of discovery right now about how much of that short cutting can we do?

23:08 The stuff I build for hobby projects, sometimes I read the code, sometimes I don't.

23:15 I just want it to work.

23:16 Sometimes I feel myself not even really wanting to read the OpenSpec design.md document.

23:22 You know, that's I'm like, come on.

23:24 You got to have some standards here, Tim.

23:26 Read the markdown.

23:28 You can do that.

23:29 But I think how much how aware we have to be of the details uh we don't know yet.

23:37 And uh I suspect that there is some early enthusiasm.

23:42 that's causing us to say, forget it, just have an agent review the code.

23:46 You gotta move faster.

23:47 You can't block on PRs.

23:49 I think that that will probably not work out well.

23:54 And when there's real money on the table, we'll find that, yeah, we still kinda do need to know something about the code.

24:01 But that takes months or years to learn that kinda thing as a community.

24:05 Yeah, that's what we're finding with security.

24:07 So they aren't this AI native startups.

24:11 are developing really fast and then they want to make some money.

24:15 They want to sell their product, their agent to the enterprise.

24:20 sometimes whatever they have written is being classified as malware or blocked.

24:25 And the bigger enterprise are saying, no, no, no, you just, you have kind of written and it kind of works, but the security's posture is really weak.

24:33 Yeah, it's just.

24:35 You got to harden it.

24:35 You got to make it compliant.

24:37 And that's when they usually come to Trenton.

24:41 They're like, I don't know what that means, that security bit, but just, just make it so that it's not classified as malware.

24:49 Yeah.

24:49 Just make it so that we can actually sell it and make some money.

24:51 Um, the, um, the question for us then it's more of a, okay, so we'll do a one-off, but what have we helped you in terms of understanding?

25:05 you know, in terms of thinking about the culture on security.

25:07 And the answers are nuanced.

25:09 mean, sometimes you don't have to have a culture on security.

25:12 Sometimes you just want to get something out there and delight your customers and make them happy with, I don't know, selling flowers on the internet or whatever people do.

25:19 And sometimes if you're doing something a bit more critical, then yes, you do need to

25:24 build that security maturity model and program within your company.

25:31 And we help you a little bit with that as well.

25:33 But the answer is nuance.

25:35 mean, it's, no two companies are the same.

25:39 But also going back to what you were saying around filtering noise.

25:45 One of the things that's super fascinating right now is that together with all the tools we have given developers

25:53 and some people are saying like, you know, we've made everybody 10x

25:56 It's probably more like 2X.

26:01 I think end-to-end is also more like 60%.

26:04 Like it's not even 2X, because there's bottlenecks elsewhere, yeah, but that's a different story.

26:10 Yeah, we're measuring it.

26:13 It's, yeah, some things are 10X, are 2X.

26:17 What's the...

26:20 Tell me a bit more, it feels, there are days when I feel like, when I remember Kafka, for example

26:27 or when I go back, there were a few kind of key voices in the community there

26:31 producing content.

26:32 I mean, you're pretty legendary there, right?

26:34 mean, setting a really high bar for everybody else.

26:39 And nowadays it feels like almost like every developer with a tool and an idea is just producing content out there.

26:48 On one hand, it's really great.

26:50 um I put some more blogs out there with what I do.

26:53 On the other hand, like, how do you deal with all the noise, right?

26:56 I mean, there is um whenever we're hiring developers and training them in our team.

27:02 How do we find that, what's the way to find that those influencers, if you will, that bring the real value?

27:11 Like, how do you filter through the noise?

27:12 How do you manage that?

27:16 What's your tips there?

27:17 I mean, I guess they can subscribe to your podcast for sure.

27:20 I mean, that's the thing you should definitely know that I've got to, you know

27:26 Substack or you just go straight to TimBerglund.com and get the goods.

27:31 That's right.

27:31 uh And it's kind of this is funny.

27:36 So everybody listening, watching, you know, and I did not coordinate this question at all.

27:41 But uh I've got a blog post I'm shipping today.

27:45 It'll be on LinkedIn, Substack, blah, blah, blah.

27:49 timberglund.com

27:51 that indirectly addresses this.

27:52 There's a part of it.

27:53 So it's kind of an expansion on my developer journey framework of discover, understand, play, build, advocate.

28:01 It's the journey that developers go through when they're adopting a technology.

28:05 There are other DevRel leaders who've got slightly different spins on that whole thing.

28:09 It's my take.

28:10 I'm not the only one who's got that.

28:12 But I think in the age of AI, as we like to say, trust becomes a lot more important.

28:19 because the marginal cost of cranking out a unit of content is close to zero.

28:24 Anybody can crap out blog posts.

28:26 Some people are making tutorial videos with AI avatars, and I really think the technology is not there yet.

28:32 I think that's terrifying in the present day, just with the quality of the AI avatars.

28:38 I think that's a super bad move.

28:40 There isn't a shortcut to building trust.

28:44 The thing is, you asked me what my tips are.

28:47 And I don't have tips because to build trust is...

28:53 You have to experience someone as trustworthy.

28:55 uh Saying, trust me, is usually what a bad guy in an action movie says at some point.

29:04 you can't just tell someone to trust you.

29:06 That's not a thing.

29:07 I mean, guess Kyle Reese said...

29:09 Come with me if you want to live in Terminator.

29:11 And that was an epic line.

29:12 And that was a very high stakes moment.

29:15 had to force, I want to say Linda Hamilton, but that wasn't the character's name.

29:20 And I had to force her to make a choice.

29:23 But trust is experienced, right?

29:24 You build trust because you experience someone as trustworthy.

29:26 And so there isn't a shortcut.

29:29 You do have to experience, we'll just say an influencer, as a trustworthy voice who doesn't just shill.

29:38 for whatever it is they're being told to shill for, uh isn't just saying the hype-iest

29:45 clickbaitiest thing, but is really trying to help a community of developers along with

29:52 some kind of frontier in their world.

29:55 And I look back at the various frontiers I've been a part of, they seem, even Kafka seems small in comparison to AI now, uh but...

30:07 Is that person a trustworthy voice?

30:09 Are they bringing value, educating, helping make sense of the world?

30:15 And you just gotta pick the people who you have that experience with.

30:19 There isn't a shortcut.

30:21 There isn't an algorithm that's gonna tell you.

30:24 There isn't this one simple trick.

30:25 uh But I think it does matter that you curate voices that can help you.

30:33 It's just like in your personal life.

30:35 You you got people that you're close with and you have something you're going through that's a big deal, some big decision.

30:40 Should I take this job?

30:41 Should I relocate?

30:42 Should my family move here?

30:44 You know, who's close to you?

30:45 Who's going to advise you?

30:46 Who have you experienced as trustworthy?

30:49 I think in the DevRel world, or more broadly, you could say the influencer world, there's that same thing.

30:55 And you got to look for those people who are in it because they actually want to

31:02 help shepherd a community across some kind of transition.

31:05 interesting because we have this exact same question asked in the security space.

31:10 People say, well, there so many startups.

31:14 What is the moat?

31:15 And I have now been through, uh I guess I worked in two startups and I've seen like the entire cycle, uh the beginning, the journey and the end, I suppose.

31:26 And especially in this area in security, the moat is, are you going to be there for the long term?

31:34 and earn the trust or are you going to be there just to do the what I call like the cherry on top only the cool part and then declare it a success and disappear.

31:46 So in many of the products we have shipped in the past really when I look back it's been

31:55 Yeah, you can't short circuit the trust.

31:57 It's something that compounds over the years.

31:59 You got to be there, help the community with, like right now we're seeing open cloud people ask about some simple questions or some even complicated questions.

32:09 But if we have some answers, we try and help and they try and help us in return.

32:15 And that part, a real moat in this year of doing things really fast is actually doing things for the longterm.

32:24 and just sticking with it and then even some of the quote unquote boring things, which um

32:34 Some of the enterprise stuff like insecurity is pretty boring.

32:37 Things like role-based access control, right?

32:40 I mean, wants to do that?

32:41 I don't want to do it, but the customer wants it,

32:44 No little kid says when I grow up, I want to do RBAC, you know, but, it's a valuable thing that people will exchange money for.

32:52 Yeah, they will.

32:53 They will.

32:54 And we've seen this uh over and over again or models that belong to the customer with the data and the learnings belonging to them so that it doesn't go to like some of the bigger

33:05 companies, right?

33:07 All of that trust is trust and all of that can be earned over time, right?

33:12 So yeah, it's interesting.

33:15 Let me put you on the spot there.

33:16 We're a startup, year one seed stage.

33:20 Give me some advice.

33:21 When, when is the right time to hire the first dev rel person or what, what are the conditions under which dev rel person for lack of a better name?

33:24 When, when is the right time to hire the first DevRel person or what, what are the conditions under which DevRel person for lack of a better name?

33:32 If you've come up with a better name, let me know by the way.

33:35 I will just say dev rel.

33:36 What are the conditions that they would be happy like in an early stage startup?

33:42 That first hire is usually what I'd call a developer advocate.

33:46 So it's the content creator speaker kind of person.

33:51 That's all right.

33:51 That's right.

33:53 No, no, no.

33:54 A lot of people say, when should I hire "a DevRel?"

33:56 And it's pet peeve of mine.

33:57 I'm like, it's not a DevRel

34:00 It's okay.

34:01 It's all right.

34:01 But a developer advocate and that person might not be, you know, three years from now when you're scaling your sales motion and you've got 300 people.

34:10 That might not be the same person who leads the team, but they're a content creator, they're deeply technical, they're a good communicator, they'll travel some, you know, so

34:20 that's the profile of the first hire.

34:22 Usually that person should come from your existing community, that's a good place

34:28 you gotta be careful if you're hiring from a customer, you know, make sure you have all the

34:32 right conversations, but...

34:33 Somebody who's already in love with your product, you've got those fans out there who is in that advocate stage of their own adoption.

34:41 They make a good, passionate developer advocate.

34:45 uh we're also assuming, like you, certainly like Trent.ai, that there's a significant developer-facing surface area.

34:55 So if you're gonna do it at the kind of A-round stage of life,

35:01 or even seed, seed is pretty early.

35:03 Usually founders are doing it then, but at the A round stage of life

35:09 You have to be sure that that developer adoption is key in your go-to-market, right?

35:15 Like you're not really going to be able to sell things if developers at your customers are bought in.

35:21 You have to know that.

35:23 uh And if you're an application, know, early days of Salesforce, no.

35:29 They don't need developer advocates.

35:30 mean, they got them now because they got APIs and they're a huge company and everything.

35:34 But it has to be that it has to be the case that even if developers don't write the check, they need to be believers.

35:43 Right.

35:43 So it's usually an A round kind of thing.

35:46 You know, in terms of a headcount, 10, 20, 30, when you're in that size, you can you could start to think about a developer advocate.

35:53 There's a delicate part of this job where nobody wakes up in the morning thinking about Trent.ai, but lots of people wake up in the morning thinking about agents.

36:06 And part of the mission, and I guess this is now just DevRel coaching for Trent.ai , and everybody gets to listen in.

36:12 uh Which is great, this is fun conversation.

36:18 Part of the mission is when we say agent, so many people think, well, Claude code or whatever

36:25 My harness or maybe uh OpenClaw, you know, they're thinking that kind of

36:32 agent that's running on my laptop or the Mac mini that I bought to run things.

36:36 uh The agents I think you're talking about are software written in the enterprise that are doing things with enterprise data.

36:44 And this, you know, this is it's a very different thing.

36:47 It's agents that developers are creating.

36:49 so part of what you have to do is now create a differentiation between

36:54 the common conception of my coding agent and agents that we're building.

36:58 You could see this play out in debates online like "MCP is stupid, all we need is CLIs."

37:05 Now that is a statement that somebody makes when their entire experience of an agent is basically Claude code.

37:13 in the enterprise is not like that.

37:14 your developer advocate, number one, has to do that.

37:17 Number two, because nobody wakes up in the morning worried about Trent, you have to think about what are the things that people wake up in the morning worried about?

37:17 Number two, because nobody wakes up in the morning worried about Trent.ai

37:20 you have to think about what are the things that people wake up in the morning worried about?

37:27 And this is developers now, not the maybe executives that you'll be selling to

37:33 but in terms of appealing to developers, what are they worried about?

37:36 And how's that connected to agent security?

37:39 What part of that total addressable developer market?

37:42 really does have an anxiety about agent security.

37:46 Maybe that's not the thing that they Googled or whatever or asked Claude about

37:51 but there is a thing that they're thinking about and the content that this developer advocate

37:56 creates, the talks that he she gives uh need to address kind of those organic concerns with Trent-based solutions.

38:07 And, you know, they're not selling, it's just that they are there to make Trent.ai famous.

38:13 uh in a way that's valuable to that developer in the audience.

38:16 So I think that's the move for that early stage DA there.

38:21 I love that.

38:22 That's a lot to think about.

38:23 um

38:24 this, so when you hire that person, we just play this back.

38:29 I love that.

38:31 All right, you know, well, I think that's a great place to land the plane.

38:35 Thanks for being my guest.

38:37 We'll talk soon.

38:40 Take care.

38:41 Thank you.

38:42 Real pleasure.