Luke: [00:00:00] From privacy concerns to limitless potential, AI is rapidly impacting our evolving society. In this new season of the Brave Technologist Podcast, we’re demystifying artificial intelligence, challenging the status quo, and empowering everyday people to embrace the digital revolution. I’m your host, Luke Moltz, VP of business operations at Brave Software, makers of the privacy respecting brave browser and search engine.

Now powering AI with the Brave search. API. You’re listening to a new episode of The Brave Technologist, and this one features Angels Menendez, who’s a cybersecurity and AI automation expert, currently at N eight N, helping enterprises scale their AI driven workflows and integrate intelligent automation into traditionally manual processes.

With a deep background in threaded intelligence, digital privacy, process orchestration, he brings a unique perspective on the evolving intersection of cybersecurity automation and responsible AI adoption. In this episode, we discussed ways that AI is changing the way [00:01:00] companies think about automation, especially at the enterprise level, cybersecurity threats.

He’s currently worried about sentiment analysis at scale, how they’re utilizing community enlisting tools to help shape marketing strategies and drive growth, and how he thinks about unlocking creativity in right brain thinkers by leveraging emergency technologies like ai. And now for this week’s episode of the Brave Technologist.

Angel, welcome to the Brave Technologist. How are you doing today? Hey, I am doing well. How about yourself? Doing well. I’ve been looking forward to this. I appreciate you making the time to come on and share with our audience. Likewise, likewise. Thank you for having me. Looking

Angel: forward to it.

Luke: Do you mind sharing a little bit about what kind of first drew you in to the intersection?

Cybersecurity and ai.

Angel: So that’s a great question. I have a pretty varied tech background. I first got into technology and networking when I was like seven years old. My father and our family moved from Puerto Rico, and originally there just [00:02:00] wasn’t a lot of economic opportunity out there in terms of jobs availability.

So my family moved to Oklahoma and my father started as a welder. And you know, during the day he’d go to a steel mill and work there. And I, I still very succinctly remember him coming home, smelling of iron. And it’s just something that just has never left me. But he started going to night school for it and computers.

This was like the early nineties. Oh,

Luke: okay. And it

Angel: was still kind of like a new thing. And you know, people were kind of like, this computer stuff isn’t gonna go anywhere, you know? Mm-hmm. But my father kind of got into it real early. Spanish was their first language. So he would bring home books like Networking for Dummies.

And I love to read. I was, I’ve always been a voracious reader, and so I would translate the books for him at the kitchen table, so like to kind of make it a little easier. So I was very young, I got into it. I thought I would be kind of the rebel though as I learned computer networking from him and from those books I was like, I never want to do this.

Like I’m an artist, not heart. [00:03:00] You know, I’m gonna, I’m gonna make art. And so as I got older, I got into web design. Mm-hmm. Letting that be my creative outlet. So that’s essentially kind of how I got into the tech world. I took a detour as a flight attendant for eight years doing web design while being a flight attendant.

So, which was, you know, quite an experience. And then from there, as I was flying around the world, I had my first child and I was like, you know, this flying thing is fun, but I don’t think I can sustain this. And I had helped a, a friend of mine get a job in the cybersecurity industry maybe 20 years prior.

It wasn’t actually cybersecurity, it was it, but he started like going up in the, in the IT world. I think of him as my mentor. We would always sync up and he would help me with my web design projects on the side. And one day he was like, you know, I, I think from what I’ve seen, like you could do cybersecurity and engineering.

He’s like, have you considered it? And I’m like, you know, I have, but I don’t have the [00:04:00] technical background in terms of an actual educational degree in this. So he’s, I’m like, do you think that, like, they’ll be okay with that? And he’s like, yeah. Like, let’s, let’s talk through, let’s, you have a portfolio already.

You have, you know, these skills. We just need to make sure you understand Linux and kind of go into it from there. So I spent maybe like six months to a year working with him and just mentoring under him and learning a lot about Linux and different backend networking stuff that I had kind of learned earlier on, but I had never been able to apply.

So I got my first job in cybersecurity, working for a company called Threat Connect, which is where he was working at, working as a customer success engineer there. And I loved it. It was just so interesting. I mean, there’s so much. So much that you can do in that space in terms of growth, like there’s really e especially when I got into it about I think six years now, there just wasn’t a lot of automation.

A lot of things were done manually, and this is kind of before AI really took [00:05:00] off. One of the big issues that I was dealing with was noise In cybersecurity. There’s just a ton of noise in terms of like, you’ll get these huge IP data sets, and the goal is really to sift through them looking for almost like breadcrumbs to see if a relevant.

Attack was actually happening or not, you know, on the customer’s backend, using tools like Sims to essentially look through log monitoring, and from there essentially extrapolate, you know, yes, you know, these ips are responsible for that. Let’s add this to a block list, put past that block list over to like network firewalls, and then ensure that.

These ips can’t get through. That was kinda where I got my start. And it was funny because the same mentor, I just remember being kind of frustrated with some of the, the tooling and my mentor was like, you know, I actually saw this tool on, it was actually Hacker News. And he’s like, it’s called N eight N.

Have you heard of it? And I’m like, no, no, I wanna take a look. So he sent me the link and I fell in love. I [00:06:00] was just like, this is so cool. I was able to self-host and coming from a web design background, I had early on learned that you can make money not just in the consulting space, but in the hosting space.

If you can get that recurring revenue, that’s even better. So I. One of the things that I had done in the web design space is I built WordPress websites and those WordPress websites I would then host. So I would make money on the design and I would make money on the hosting, which was very nice to have like a recurring income.

Luke: Mm-hmm.

Angel: And I remember thinking when I saw N eight N for the first time, like, oh my God, this is like. The, the WordPress of automation. So I was able to kind of internally at that cybersecurity firm I was working at build like my first real like in production workflow, which it ended up saving the company.

I think it was like 50 man hours. Per customer, per customer success manager per month. Wow. And it, there was [00:07:00] like, my goodness, I think there was like 15 to 20 of us in the customer success side. So the amount of hours it saved was huge. And I think like. Four years later, I was working for another cybersecurity firm and we purchased that product Threat Connect and they sent us the N eight N link.

And I’m like, I built this, I’ve built this. We’ve been here years and they’re still using it. This is great. I, I left the legacy, you know, so. That’s awesome. That’s awesome. Yeah, it was very, very cool. So. The way this all kind of connects back to ai, at my most recent organization in cybersecurity, I was working at Palo Alto Networks kind of doing something similar.

So I jumped around, you know, going from a cybersecurity firm to firm and I was working at Palo Alto Networks and that was when ai, Chad, GBT kind of was first released. And I was just obsessed. I was like trying to find ways to hack it. And because I had the ability to work. On the API side with N eight n in such a visual, easy to access way early, early on, I was playing with the system prompts and trying to [00:08:00] build ways to get around, like a lot of the monitoring and, and tooling that they have in place on the chat GPT interface in order to like see what I could just do with it.

Like mm-hmm. Just what, what can I do? What can I play, what can I generate with this? Because I was also very involved in the N eight D. Even though I didn’t work there, I spent a lot of time in the community forums. Essentially what I would do, I had like a a little trick. I would essentially answer a lot of the easy questions to force the employees to answer my complicated questions For me, now, I’ve answered my 40 questions.

Get to work, you know? So it worked out great, but then I kind of made a name for myself that way. And one of the marketing heads reached out to me while I was working at Palo Alto saying, Hey, you know, would you be interested in working for us? Cybersecurity is kinda a big industry for using low-code automation tools again.

To reduce noise. So that’s kind of how they reached out. I’m so excited. The opportunity, again, having built a lot of my own personal automations at home using N eight [00:09:00] N, it was a really easy shift in, I don’t wanna say careers ‘cause it was very similar career-wise, but more like a shift in jobs. And I really enjoyed it.

And one of the first ways that like I was able to kind of have an impact is seeing just how powerful AI can be in reducing noise. So, again, kind of in cybersecurity, one of the main issues again is like sifting through mountains of noise and being able to essentially isolate noise that essentially is relevant to the situation and noise that is just background noise that we wanna eliminate and, you know, go from there.

So. Basically, one of my goals was like, you know, how can we utilize these tools, this, this platform to make that easier? So I had, I had built quite a few cybersecurity workflows that we published on our gallery. One of the ones that I’m most proud of was actually not noise reduction, but automated sim classification of alerts.

So that’s [00:10:00] another big one, is like in cybersecurity. Typically you’ll kind of have different layers of workers. So one of the early layers is analysts. You’ll have analysts where data comes in, they analyze the data, and from there, you know, kind of escalate it, kind of go from there. The problem is that like many analysts.

Not all, of course, but there are many analysts that are very new to the industry. Cybersecurity itself is very kind of young. Mm-hmm. And so it makes it difficult for, for onboarding new analysts to have to like teach them the basics of the industry. Teach ’em like for example, something like Mitre attack, which is like a catalog of different hacking attempts and triage and how to like identify whether this noise is coming in is really a hacking attempt or not.

And so one of the early builds that I built in N eight N and released a video of was a way to connect Mitre attack to your AI using a vector store. So essentially taking that catalog of information, [00:11:00] vectorizing it, and then attaching it to your AI agent so that it can use that to essentially. Get relevant context.

For example, you could connect it to a sim and those alerts first go through that AI agent, it identifies what, potentially, what kind of attack it could be. Um, it tags it with different tactics and procedures. And from there, it essentially assigns it to a, an analyst via tool like Jira, for example. Mm-hmm.

Mm-hmm. And so those kinds of. Intersections for me are really interesting because it doesn’t exactly take away jobs so much as it helps give newer workers to an industry. It equalizes the playing field by giving them more context to be able to do their job more effectively. That to me, was kind of where that intersection came from, and a lot of what I do now is mostly kind of an intersection of the marketing side as well as it ops and cybersecurity essentially.

Helping our internal teams better understand where processes can be improved, as well as talking with some of our [00:12:00] customers, both, you know, sales and customer success in helping them implement existing workflows to be able to automate some of those more tedious, you know, cybersecurity and automation related tasks.

So I hope that wasn’t too long-winded, but does that No, no, no. That’s why you’re

Luke: here, man. I think this is great. It is very underrated. The path of entry from web designing and development into the security world. And I feel like security is one of those areas where. There’s so many things that happen like at the front end level, and there’s so many facets to it.

I want the audience to understand you don’t have to go get a comp side degree to like get into security. Like you see how much records there is in everyday web development and seeing where you can kind of patch the holes there. I think it is a really cool point you made about how AI and agents can help you sift through the noise.

From a practical security perspective, what are the new and emerging threats that you hadn’t had to really face before that you’re having to think about [00:13:00] more now than in the past?

Angel: Ooh, that’s a good question. I would say so lack of understanding of the tooling in general I think is kind of a big issue.

I feel like. We have these, these powerful AI tools at our disposal, but I feel like many people in the industry don’t quite know how they work. They’re kind of like a black box.

Luke: Mm-hmm.

Angel: And I think that if you don’t fully understand. Things as simple as like file structure and file management. Like you have to kind of start at the bottom and work your way up.

If you kind of start with ai, it makes it very difficult to have an understanding of just how useful it can be. Mm-hmm. So I think I would say that like in a situation like that, the biggest like kind of emerging things that I’m seeing on my end is kind of inexperience coupled with a lot of.

Responsibility and power and not knowing exactly how to implement some of these tools, especially on the [00:14:00] AI side in order to essentially. Maximize the output, but also ensure that the data that is being sent to some of these models is protected or mm-hmm. You know, not exposed in many. Mm-hmm. In many ways.

So like, one of the things that I would see kind of early on is passing text or URLs directly to services like virus total and. Unfortunately, like that exposes a lot of this data because some of these databases are, you know, public essentially. Mm-hmm. And some of those similar issues kind of have transferred their way into AI in terms of like, you don’t control some of these models.

Like you don’t control, like chat GPT or open AI’s models. And you need to be very careful about what you’re sending into these models and be careful that they’re not being used as part of the larger training data sets. Being able to like build around some of those. Security vulnerabilities. I guess you could think of it as like, it’s very important, [00:15:00] but again, it kind of comes back to having kind of a baseline understanding of what is happening on the backend, on the API side on the data side, and making sure that you understand kind of from beginning to end where your data is currently living, where it’s going, is it on the cloud, is it local?

Where are these models living? Like are you using a cloud model? Are you running something locally like Alama? Essentially being able to decide what the best tool for the job is, you know? Mm-hmm. Sometimes you do want to use tools like Open ai, but there’s many times when you want to self-host and you want to run these models locally.

That’s one of, I think the benefits of using N eight N is, and one of the reasons I got so excited is I want to be as privacy focused as I can, and I hate to expose data when I don’t have to. So if I can control the infrastructure, I feel a lot safer. In terms of like, you know, and, and you’re still kind of exposing yourself now you have to be the network admin, you have to be the, you [00:16:00] know, your own CISO in many ways.

Yeah. Yeah. But if you have a firm understanding of like, you know, exposure in terms of like what courts you’re exposing, what’s open to the internet, what’s not, air gapping, things like that, you can really protect yourself in ways that you normally couldn’t do if you didn’t self-host.

Luke: Can you break down a little bit what N eight N does for customers and then for the end user too?

Angel: Absolutely. Yeah. Great question. So N eight N is an automation workflow platform that allows you to essentially build out. I like to think of it as code with blocks in many ways. Mm-hmm. This is something that my children, when you ask them like, what do you do dad? Like, their answer is like, well, dad plays with blocks all day, obviously.

Mm. You know, they’re, they’re three and six, so they kind of simplified a little bit, but in many ways what you’re doing is you’re abstracting a way, a lot of coding to kind of this diagramming flow building. It’s one of the things that I really appreciate about the platform because. For example, when I worked at Palo Alto [00:17:00] Networks, we had, usually when we were building out flows for our customers, it would take two employees to do it.

So you would have your customer success architect, which would come up with the general idea, and they would have a, a very good understanding of all the different APIs involved. They would create a document or a, a diagram. To essentially show what steps would need to be taken, and that would become our blueprint, but then would be handed down to us in the engineering side to essentially deploy the actual code, usually Python or something like that, internally.

And one of the benefits of N eight N is that you can reduce that down to one step because you are doing both the diagramming and the coding in that same step, which I, I think has. Seen a big rise in the prevalence of like vibe coding in terms of like just sitting down and like building out these complex flows because to allude to the name, like the blocks are called nodes, so you’re able to see the inputs and outputs kind of mm-hmm.

[00:18:00] Within those nodes. And being able to do that gives you this very linear approach to coding that you can store in your head in a visual way. One of the things that separates N eight N from companies like Make or Zapier, is the ability to essentially self-host and own your own code. So kind of going back to that WordPress analogy, being able to own your own code is huge because.

In many ways, in this day and age, it’s very easy for the companies that control the cloud to control the cost. Mm-hmm. And many times, like it’s arbitrary and it could be at a whim and you know, you may not like a price increase, but you just gotta suck it up. And for. You know, small business owners or medium sized business owners, those kinds of costs add up very quickly and it becomes very difficult to either scale or to have an idea of what your expenses are at a given, you know, at any given time because you’re not able to, to set a stable cost, right?

And by [00:19:00] being able to self-host, you’re basically, you need to invest the knowledge on how to deploy. At the same time, you lower your cost to almost the cost of the electricity of running a server, and that allows you to keep your pricing very stable. It allows you to own your own code. In terms of, like, one of the things that I like about N eight N is that the, the code is stored as JSON.

So you can save that, you know, anywhere. You can, you know, copy and paste it into your canvas and you can just deploy as you go. So. In a nutshell, the platform itself is about automation, but the way that you, that you host it, I think is one of the biggest kind of differences between us and our competitors.

And in addition to that, the UI is very, very intuitive. So everything is kind of. Inputs and outputs. The name of the company itself, the N eight N, is short for Node di, so it’s kind of like automation, but with nodes. The N eight NI think comes from, I think Kubernetes, kind of that K eight s or Ks, you [00:20:00] know, kind of breakdown.

So it’s kind of similar in that way and it’s kind of stuck, but it’s kind of given us this ability to give customers a choice. In terms of like, how do you wanna deploy this? Do you want to go with our cloud where we manage the infrastructure and kind of take that stress off of you? Or you know, are you a cybersecurity firm that has a lot of limitations in how you can host?

Like do you wanna self-host? Do you have a lot of PII flowing through your automations? Do you need to like air gap that and just have that running locally? You can do that. I feel very, very privileged to be able to work for an organization like that that has so much flexibility and I think that’s what’s led to some of our like massive growth that we’ve been experiencing lately.

Luke: That’s awesome. Toward the end there, you brought up a really good point on the self-hosting piece that it’s something we kind of deal with at Brave a lot, where we think about first party relationships and third parties, and it seems like the ability for a business to self-host easily. With the way that data providence and [00:21:00] user information flows and breaches and all that, one of the things I’m really kind of concerned about right now is just that there’s this balance of convenience with these prompts.

It’s disarming, right, like for the user where you just stop thinking about certain barriers that you naturally put up with giving information away. I thought ad tech was bad. You don’t know where these models are sending anything. Oh yeah. It was something we dealt with a lot where we were, especially when we brought like advertising, you know, that was privacy preserving to the default when we would talk to customers and it was like, hey, like maybe there’s like one or two things where you can actually self-host it and have it work like a Google Analytics.

But on the AI side, it seems like there aren’t really a ton of either here, people talk about local models and stuff like that. But not a lot about the self-hosting. So it seems like really interesting from a user first perspective on things as an option for businesses. So that’s super cool to hear. Are there a lot of customers that are self-hosting right now?

Is that a major thing you guys try to do or is it just pretty [00:22:00] new?

Angel: Yeah, I, you know, that’s a great question. I need to look into, I know we have a lot of like internal metrics on that, but yeah, I’d say a very large majority of our customers do self-host. I think it’s almost like. A gateway drug anyways, you start running it at home for personal use and then you start seeing the power of it.

Mm-hmm. And then, you know, you go to work and all of a sudden, you know, you start seeing like flows in your head or in your dreams and you’re like, man, I’m doing this manual process at work that I could like automate using N eight N. And then you start talking to your internal teams and you start deciding on infrastructure.

And if your organization doesn’t have like. A robust DevOps team to be able to like self-host. Then you kind of then graduate to the next level, which is like, hey, well then let’s move into n Aden’s cloud and let’s use their, their own cloud product in order to like power this. But what I found is that in terms of like experimentation and MVPs, most.

Serious automators will start with [00:23:00] self-hosting and then kind of work their way up as they need to. I’d have to look internally to see, you know, what the split is between hosted and Unhosted clients, but it is definitely a non-trivial part of our organization’s community. No, it’s awesome.

Luke: I mean, even that bit is impactful, right?

Sometimes it, people would think about it like, oh, it’s just like something outta my garage or something. It’s like a science project versus like something that serious businesses are using. So it’s super cool to hear that. It also seems like right now compared to other types of emerging tech, AI is one of those things where there’s so much hype and you’ve got like Fortune 2000 companies, executive teams kind of pushing mandates across the organizations for businesses that.

Are established but haven’t done AI to like put AI everywhere. And then you’ve got also on the other end, like this flood of tooling and new models and different types of models that are doing different types of things. Like how do you balance that lack of market fit with pace and tempo of new entrants and, and all that [00:24:00] you have to learn there with providing good security for users and for businesses.

Angel: Again, I think it comes back to the users. Own knowledge of data management in many ways. Mm-hmm. And understanding what surfaces is. Is this data touching, is this touching like a remote server? Is this touching an internal server? I will say that internally at N eight end, one of the focuses, and one, I think one of the ways that the platform is powerful and kind of one of the differences between us and many of our competitors is our focus on standards.

Being able to essentially not build. Custom kind of bespoke connections, like using these standards to let users decide what they want to do. For our AI agents, for example, I think one of the reasons that we were very successful earlier on is not tying ourselves to one specific AI agent or model or tool, but using standards to allow for the [00:25:00] integration of just about.

Any AI model or large language model to be integrated into your workflows. And because we were kind of an automation first company to begin with, we already dealt with a lot of standards in terms of like, for example. Using JSON for? Mm-hmm. Just about everything, like internally from data management to workflow storage, using things like Curl for API connections, you know, being able to like go to just about any API documentation page, copy the curl request, and import that into the HGTP request tool.

Makes it like a Swiss Army knife, essentially. Mm-hmm. I’d say the answer here is. By focusing on standards and not so much on the one-off solution, we’re able to kind of cater to a larger audience. And what we found is that that audience, for the most part, tends to know what they need and want. And it’s mostly just getting, giving them that availability to be able to integrate using.

You know, whatever standard works best for [00:26:00] them to be able to deploy either the AI models or their automations or, you know, whichever one works best for them. And not worrying too much about, you know, having the, the perfect answer. It’s more just like finding the best standard that goes with the platform.

Luke: It’s a great point. I’m personally haven’t dug as far into this as I probably should have, but how are you seeing kind of standards bodies adapting to this wave of ai?

Angel: I almost think of us as an organization as having like kind of two different kind of focuses, right? And they’ve just converged very naturally.

Mm-hmm. Like the automation side and the AI side have converged into what the product is today. To answer your question, you know, what I’ve seen is that the standards are being built by the biggest players. So like you’ll have open AI for example, like early on, built kind of a standard on how model communication kind of takes place.

We’re able to reuse that standard across many different large language models. [00:27:00] You can actually integrate some models that we don’t have need support for into N eight N by utilizing those standards that kind of open ai. Pioneered. So what I would say is kind of a mix like for kind of like the automation side.

There’s so many standards that have paved the way for what we’re seeing now in terms of the AI side, but at the same time on the AI side, we’re seeing, because it’s kind of like the Wild West, we’re seeing them create these new standards. They kind of duke it out and. The easiest standard to integrate ends up winning and starts moving forward.

We’re seeing that kind of, I think on the anthropic side with MCP as well, being able to like deploy these, these standards allows for much faster, builds much faster integrations. So again, like I think that that has allowed us as an organization to grow by kind of piggybacking on these standards, but at the same time, we have to be very picky.

In terms of like what standards we’re gonna go with, because a [00:28:00] standard that you know, might be hot and in fashion today might just be a bubble, you know? Right. Oh, totally. Yeah. Yeah. Where are you? You know, you have to like reverse track and you wanna avoid that where you can. I’d say like it’s really a mix.

Like it’s knowing what has worked in the past. And looking at these emerging trends and trying to forecast for the future and trying to build. So one of the things we do internally and, and the big projects that we’re working on internally and that I’m a part of is sentiment analysis on a bulk scale.

We have access to all this data online from places like Reddit or Twitter or all of these, like social media companies have created this fire hose of sentiment that is just ripe to be analyzed. So. By using Innate Ends ability to, to scrape and pull and, and merge tools together. We can build internal tools to try to get a pulse on, you know, what the Internet’s talking about, what the Internet’s excited [00:29:00] about, and from there be able to decide, okay, you know, we’re seeing a lot of noise and it’s sustained over, you know, X number of months in this particular area.

This is probably where our product team should focus on. This is where we should move forward. And I feel like that has been. Part of our success has been to utilize the community that already exists around our product and all, all these AI products out there listening to them, which in the past without AI, would’ve been very difficult.

Like yeah, it would’ve been like I can, I can only imagine like the flows and builds to try to standardize the outputs. So many. If checks, you

Luke: know,

Angel: if you know key words exist now we can really kind of boil down whole conversations down to like excitement or fear, apprehension, or these tags that allow us to then move forward in a way that builds more confidence.

It feels more data driven. But you still have to apply a lot of knowledge in terms of like, what channels are we listening to? Why are we listening to these [00:30:00] channels? Are these channels a reflection of the greater a community or are they hubs for a bubble to burst? Mm-hmm. You know, so like you have to be very careful.

But I think what AI has done is it’s given us, given us the ability to listen to our community at scale, and then from there make data-driven decisions, which I think is one of the things I’m most proud of with N eight N is. We try to make as many decisions as we can based on data. I, I’ve worked at some organizations in the past where product features were driven by, you know, the least common denominator, which is like, how much money are we gonna get by implementing this, this, you know, feature or tool.

And at N eight N, what I’ve seen, because so much of our success has been driven by our community, what we have learned, you know, early on is by listening to that community. We can keep growing because the community is very vocal. They’ll tell you what they want. And so by listening, I feel like we’re able to outpace our competitors in terms of like, this [00:31:00] might be a fad right now, but this is what our community’s saying.

Let’s go down this path. As opposed to like, this one organization is gonna pay big bucks for this one feature. We’re gonna build like entire, you know, workflows around that. Instead, it’s like, what can we do to help the greatest number of people?

Luke: Yeah, a lot of it’s like listening into a mirror for what we do too.

I mean, the community element is huge, right? Especially with building software and like kind of open software. If you’re not listening to those people, you’re really just kind of running with blindfold on basically because they’re gonna be with you when you’re in ups and downs. And there’s that kind of like volunteer element of like contribution that the communities have in in software development.

See, you see it in other places, but like it’s not a common thread in the commercial space at all. And so it’s super cool to like hear that. And my mind is kind of going to this place as we’re talking here too, that I think is really interesting where there’s so much fear from people all over the place.

You hear people talking, oh, you can have a [00:32:00] 10 x. Coder or, you know, AI is gonna do all of the work for us, or my wife’s a product marketer and they all were like using prompts like super early and they just kind of organically did it. And they’re like, wow. But you’ve got this fear around the plow and, and the tractor.

But what I’m hearing with what you’re telling me and almost every different pillar of this discussion, it seems like NAN is kinda one of those things where you’re not gonna get rid of the engineer or the coder or the architect. You’re really gonna give them building blocks to like do really cool things.

Their job’s not gonna go away. It’s just gonna change because so much of it doesn’t have to be manual anymore. Is that you guys are kind of helping to build some of the power tools that aren’t just the robot ticking over everything, but like really like what’s the edge that we’re gonna get? Like, ‘cause that sort of just keep hearing.

‘cause even on that last point, right, like with community where. We have people that are out there kind of, it’s almost like you’ve got layers that that information filters up through and, and trust and all of that. And like, it seems like what you guys are doing is kind of helping to like [00:33:00] optimize that too and rate and, and get quality as fast as possible.

You guys might be doing the Lego blocks now. Where do you see this going over the next five years or so?

Angel: And I wanna build on a point that you just Yeah, yeah, sure, sure. Please. Yeah. I feel like this isn’t the first time this has happened, like mm-hmm. In, in different spaces, right? Like, I feel like early on, one of the big fears when I was in web design was.

Oh my gosh. You know, we have Wix now. We have all tools that are gonna put web developers out of business because why would you pay someone so much money when you can just go on Wix and throw together, you know, a five page website with just a few clicks?

Luke: Mm-hmm.

Angel: And what I learned is that I think it comes down to language.

It’s very difficult for an end user to communicate. To in ai, to a web design website, like exactly what they want because they don’t know what questions to ask.

Luke: Mm-hmm.

Angel: And so I think what’s happening [00:34:00] is kind of similar to what happened kind of in the WIC space, is yes, some of kind of the low effort, low hanging fruit has been kind of trimmed off.

But the, the web developers that understood the power and were able to utilize tools like Wix or WordPress were able to see, okay, this is just another tool in my tool belt. You know, if I’m building a brochure website for a family friend, I can build something in Wix in five minutes and get them off my back much faster than if it’s a paid customer that needs, you know, user management and CMS or CRM capabilities.

At that point, I’ll use WordPress and instead of like getting angry at an emergent technology, you, you see, how can I add this to my tool set? And I feel like that’s exactly what’s happening with N eight N. It’s essentially. AI is yet another node in our tool set. The the company is still automation first.

It’s still essentially how can we reduce the pain points that we’re experiencing? And a big part of my job, I work with enterprises to help ensure that [00:35:00] they can scale effectively. And a big part of that is not taking away jobs. A big part of that is reducing friction, because in larger organizations, what we’re seeing is siloing effect of.

Teams being entrenched in one tool, one platform, one website to do their work, but they don’t have coherent systems to offload that data that they have been massaging or working on to the next department. So you might have sales using Salesforce, and then you’ll have, you know, customer success using their CRM.

And then from there you’ll have the marketing team using Notion and it’s like, okay, who’s gonna step back? Look at these, these tools, not as the end all be all, because some of these departments are so entrenched that that’s their entire world. But how do the APIs and backends allow the data from the Salesforce side to then push a web hook into the customer success side?

So you have a data packet with all the information needed to push it into the CRM, and then how do we feed that data from the CRM as graphs and charts [00:36:00] for the notion side? In terms of the marketing to then say, Hey, this is the, the patterns we’re detecting across our customers, how can we market more effectively to them?

So not to kind of go on a tangent, but to answer No, no, this is great. Essentially, like where I see NAN going is kind of similarly utilizing AI as yet another tool in the tool belt, right? There are many, I think like automation engineers. That essentially are gonna continue to use the visualization and building, you know, OneNote at a time because they think in that linear path.

Mm-hmm. But some of the stuff that we’re working on internally is the ability to generate workflows from a prompt. So we’re able to type in, you know, I need to build a workflow that connects, for example, you know, Salesforce to this, you know, bespoke CRM that we’ve built, that’s headless to essentially transfer the data seamlessly between this and that.

Right, and you type that in, you hit enter. The AI takes that, it looks at the documentation for all of our [00:37:00] nodes. It takes a, you know, it looks at all the templates that have been, you know, submitted via our own community and using those patterns that it’s really, you know, that’s, I think where AI is really the most powerful.

It’s finding these patterns that we can’t really see as well, merging those into an output that is. A fully finished workflow and then you know, with a few clicks you have an output. You have a bunch of nodes all put together, but I think the difference is. The problem with AI currently in my opinion, is that it’s still a black box.

You can do some amazing things with ai. You can have it build a Python code that will do all of what you know, N eight N is doing in a particular flow. The problem is. How are you gonna debug that when it starts? When, when there’s a change, for example, in the backend, API that it’s connected to, you need to make sure at that point that you, you either understand the API to fix it yourself.

You need to be able to communicate. And that’s where it comes back to asking the right questions, knowing what questions to ask. It’s like, Hey, you know, the API is changed. Like this has been deprecated. We [00:38:00] need to support a different endpoint. This is like, you know how the data needs to be structured to be able to support it.

How do you communicate that to the ai? And I think that is where NAN shines because it takes away that black box effect. You’re not just talking to rock code. You’re able to see the data flowing through the blocks and seeing the input and output of each and being able to say. You know, with just a little bit of work saying, oh my gosh, here’s the error.

The problem is the API, we just need to tweak, you know, this, this here. Boom, we can fix it and now we’re back in production. So I think like the ability to visualize, and I was talking to one of my colleagues the other day about this, and I think that N eight N is uniquely positioned to allow creatives.

To code, and I think that you have kind of this right brain, left brain split, right? So you might have your left brain engineers, which are very, you know, logic driven and can code, but don’t have the imagination to come up with flows without having an [00:39:00] architect step in and say, Hey, this is what we need.

Can you build it out? Instead, you’re taking the creatives, they have all the ideas. But none of the technical expertise. And we’re empowering them to essentially say, Hey, here are the tools. This is what the engineers are using on the backend. Abstracted away to a simple input and output go crazy. What can you build with this?

And I think that unlocking the creativity of those of us that are more right-brained gives this like superpower to allow. Users to be able to build stuff that we’ve never seen before, like mm-hmm. Building, like granted, some of those things aren’t always the most optimal. Or many times you’ll use AI in ways that you’re just like, you know, you could use a math function for that.

You don’t need an AI doing math. Like there’s not, but it works. You know, at least you know, for the MVP. Yeah, and I think unleashing that creativity is really important because yes, I think that AI is going to probably take away kind of the low hanging fruit, but it’s going to give these new superpowers to this [00:40:00] whole new segment of users that I think is going to be a net positive effect in terms of creativity, in terms of output, in terms of giving solutions that weren’t really either possible before or able to be imagined before.

Does that kind

Luke: of answer the question? Yeah, no, that’s fantastic. I think it’s really, really frames it up well. Where can people find your work or, or reach out or see what you’re posting or talking about online? Do you, if you, if you do that.

Angel: Yeah, absolutely. So yeah, you can find me on, on the most of the socials under DJ Angel.

And I’m not a dj per se. Typically, DJs I think is disc jockey. I consider myself a digital jockey. I like to think that, you know, I know how to, how to play the internet waves. You know, you can find me on like x.com/dj angelic. You can find me on LinkedIn. I’m, I’m spending a lot of time on LinkedIn. I think it’s under linkedin.com/angel g Menendez.

That’s a great way to reach out to me and connect and like ask questions. Mostly on our YouTube [00:41:00] page on N eight N. Like a lot of what I do is I work with my colleague Max to build content to help. In my case, my goal is to build content to help enterprises and to help kind of make their lives a little easier.

And Max works more on the the larger community and kind of the more creative ideas. So we work together to try to create content that is entertaining and hopefully educational.

Luke: Awesome, man. Well, angel, I, I really appreciate you taking the time. Love to have you back to you to check back in on, on how things are going.

Thanks for coming by today. Really appreciate it.

Angel: My pleasure. My pleasure. Happy to be here.

Luke: Thank you for having me. Thanks for listening to The Brave Technologist Podcast. To never miss an episode, make sure you hit follow in your podcast app. If you haven’t already made the switch to the Brave Browser, you can download it for free today@brave.com and start using Brave Search, which enables you to search the web privately.

Brave also shields you from the ads trackers and other creepy stuff following you across the web.