Software Spotlight

Michael Bernzweig (00:01.936)
Are you ready to unlock the secrets of B2B SaaS success? Tired of sifting through endless noise to find actionable insights that actually move the needle?

Welcome to Software Spotlight, your weekly deep dive into the transformative world of Software-as-a-Service. Your host is Michael Bernzweig, who in 1998 launched Software Oasis as one of the first platforms enabling businesses to download, license, and deploy software instantly across their networks with a single click. Today, Software Oasis has evolved into one of the leading communities where businesses find top tech consultants across the USA and Canada.

Each week, Michael sits down with industry titans, innovative leaders, and game-changing executives to bring you exclusive insights you won't find anywhere else. From emerging technologies to market trends, Software Spotlight delivers the strategic intelligence you need to accelerate your SaaS career and stay ahead of the curve.

Join our growing community of tech professionals and decision-makers. Subscribe now on your favorite podcast platform and visit softwareoasis.com/subscribe to get our bi-weekly newsletter delivered straight to your inbox.

Get ready for weekly data, trends, analysis, interviews, and insights that will propel your B2B SaaS career forward.

Subscribe to our bi-weekly newsletter https://softwareoasis.com/subscribe/ to stay updated with more insights from technology leaders and transformation experts.

I'd like to welcome everyone to this week's edition of the Software Spotlight. I'm your host, Michael Bernzweig. And joining us from Undetectable AI is Christian Perry. He's the founder and CEO of Undetectable AI. And we have some exciting topics to get into today. So with that, Christian, welcome to the podcast.

Christian (00:26.51)
Thanks for having me on Michael, appreciate it.

Michael Bernzweig (00:28.218)
Yeah, now wonderful. really appreciate you taking the time out to join us today. And maybe for anyone that's listening that may not be familiar with undetectable AI, can you give us a little bit of the backstory on both your journey getting to where you are today and a little bit about what the company is all about?

Christian (00:50.222)
Absolutely. So I've always been a serial entrepreneur. This is not my first rodeo. My last company was called Chatterquant in the financial technology space. That's kind of where I started dabbling in artificial intelligence. We did natural language processing of social media. So essentially anytime somebody posted on Reddit, Twitter, Instagram, anything like that, we ingested that information, identified, you know, what stocks were they talking about? What price targets were they talking about? How are they discussing the products?

and packaged that up into Tangible Insights and sold that off to hedge funds, banks, and traders. That company was acquired, transitioned into Undetectable AI as LLM started picking up two or three years ago. And the rest is history there. Undetectable AI takes AI-generated content from things like ChatGPT or any of these large language models and kind of...

Michael Bernzweig (01:35.527)
Congratulations.

Christian (01:47.022)
takes the AI out of it. So we call it humanizing that content, make the content look less like it was generated by chat GBT while also making it undetectable so that you can go and use AI and basically arm the little guy with AI machines without the penalties of AI. We can talk a little bit more about what those penalties are later on too.

Michael Bernzweig (02:09.198)
Sure. And there's no more interesting topic to our audience than AI because I know there's, it's a moving target. Things keep transitioning and changing every day. And there's a lot of, lot of details that people are wondering. So I guess that's a great place to start. So as far as, you know, even the company name, Undetectable AI, how did, how did you get to the name?

Christian (02:37.678)
Well in terms of finding this name It's the same struggle. I've had with anything. I've tried to build it's trying to find an available domain, right? You know, I might think I have the perfect name But if the domain is not available, then that's not what I'm gonna build Or what I'm gonna name my product and so it was you know trying to understand What are we trying to create at it at its core and it was undetectable AI and undetectable dot AI's domain was still available and

Michael Bernzweig (02:52.836)
Sure.

Christian (03:06.956)
That's how we picked that.

Michael Bernzweig (03:09.61)
Fantastic and it just says exactly what you do and what you're all about. I think how has it become such a question mark and is it a valid question for companies to be concerned with, you know, detectability within their content?

Christian (03:27.564)
Very good question. That all depends on your use case for using a large language model. If you were to ask me, and I would say our whole team believes people shouldn't be stifling AI, we shouldn't be stopping putting it on different platforms, but the reality is a lot of these platforms are cracking down on AI generated content. And it does make sense a little bit from their perspective as well. Facebook, for instance, if

Michael Bernzweig (03:54.749)
is to have, for instance, if we are able to have foreign US people recognize who's in the local place and who's not in each other's house, and then severely de-oguish the university. Which is really nice to be able to have foreign foreign students.

Christian (03:56.546)
their whole platform just becomes AI's posting all over the place and AI's liking each other's content that severely devalues the user experience on their platform, which is why some of these platforms are cracking down. And examples of that are Google. Google's cracking down on if you put out too much AI-generated content on your website.

Amazon's cracking down on AI generated ebooks and is even discussing doing so for product listings. We have email providers using that as one of the metrics to determine whether your email goes in the spam folder. Obviously not the only metric. If I've been emailing with you, Michael, for months, it's not gonna, if I have Chad GPT rewrite one of my pieces of content, it's not gonna toss that in the spam folder. But if you're doing bulk outreach, you're more likely to get put in that spam box.

Michael Bernzweig (04:28.156)
Right.

Christian (04:43.874)
with AI-generated content. And again, that's all about protecting their platform and the user experience on that platform.

Michael Bernzweig (04:51.313)
Yeah, and I think if you look at it from that viewpoint, that lens, think, know, chat GPT, perplexity, Gemini, and even some of these newer search engines. Obviously DeepSeek has been all over the news as of late. A lot of these search engines are just trying to surface unique and genuine content that...

adds to the internet versus regurgitating what's already out there. And that's obviously one of the reasons the podcast is so important to us here at Software Oasis is we're trying to do just that, trying to generate unique and content that's adding to the internet.

Christian (05:37.176)
Sure, yeah, that makes a ton of sense and that's what we see with Google as well. And one of the reasons why they're using these detection metrics is for exactly that reason, that a lot of what we're getting with AI is not unique insights, but regurgitated content based on what those models are trained on.

Michael Bernzweig (05:55.951)
Yeah, and I think that that's a great, great detail there. One of the things that has, you know, is just look at some stats recently, and it's got to have the folks over at Google shaking in their boots because I read that over this next year, the volume of searches that begin with traditional search engines are on track to decrease by 25%.

Whereas the volume of searches starting with searches from AI are on the increase. So it's a very big shift in the market, a lot of changes.

Christian (06:36.3)
Yeah, no, absolutely. And we see Google responding to that. When you search now, you're a lot more likely. And I just saw something about how the complexity of the topics that Google is starting to answer with their AI summaries has increased. I don't want to give a number because I don't want to misquote, but significantly increased.

how many complex searches are being answered by AI rather than, hey, you know, the traditional here are all the links and go find the answer within the piece of content that you're, that's being served to you. It's a huge concern for them.

And also going to be pretty impactful to all the businesses that rely on that search traffic to generate their income. For instance, if I have a complex legal issue that I'm trying to find the answer of, how would I find that in the past? I'd do a Google search. I'd click on the link and it would be from XYZ law firm. XYZ law firm would give me the answer to my question and say, hey, if you have further questions, feel free and call us.

Schedule consultation. I've done that multiple times right if I get that answered by AI I'm not going to be led to XYZ law firm and or I'll just enter follow-up questions and refine my query and Bypass hiring the law firm altogether

Michael Bernzweig (08:00.027)
Yeah, and it is absolutely right on target. And you know, it's interesting, AI searches are on target to grow by 35 % per year over the next three or four years, which is just a crazy stat. And we've actually, on this end, just finished, from our end, you know, we have done tons of analysis internally here over the last couple of years to figure out

what is it that it takes to rank in AI? And we've just put together a course from our end that covers all of the details and learnings that we've put together over the last couple of years. So a lot of people are just starting to delve into this area. And the number one question I get, I mean, obviously I started on the internet back in 1998. the...

early days of the internet back in September of 98, which is when Google started. So I've seen a lot of changes over all the years. And for a lot of friends, a lot of family, different people I speak to, they're like, hey, Michael, can you spend a little time helping us get ranked online? And it's like, it's not what I do. So this course was kind of the culmination of about a year's worth of effort with several hundred.

Christian (09:14.168)
Sure.

Michael Bernzweig (09:21.614)
videos and seven weeks of learning to help people kind of figure it out on their own, but at the end, then hopefully make my phone stop ringing. But at the end of the day, as far as undetectable AI, are some of the, I don't know, like if you were to say some of the concerning implications of AI content that the businesses and society in general are doing, like what's going through people's heads? are they wondering about?

Christian (09:52.046)
So is that the question is specifically about where are we going with AI and what are kind of the some of the repercussions of that?

Michael Bernzweig (09:56.719)
Yeah, what are the topics that you hear people concerned about that they're looking to undetectable AI to help them overcome? Why are they, the users that are using the platform, what is it that they're trying to achieve?

Christian (10:15.502)
Absolutely. So when I use something like chat GPT or I put in a prompt, I mean, one of my biggest use cases for a chat GPT is I just Jot out. I'm not writing a really fancy email. I just kind of jot out my bullet points of what I'd like to address in that email or put a really dirty and sloppy, you know, version of what I'm going to email out into chat GPT and ask it to rewrite it. Um, when I do that,

I get a lot of things that very much look like it's AI generated. For instance, in a business email, it goes dear, which I don't think is a super appropriate way to reach out to a business contact. Maybe some people do. It's not how I do it. And so same thing with best regards, it signs kind of the signature the same way every single time and doesn't really look like, it looks AI, right? And it's because it is AI.

Michael Bernzweig (11:05.656)
Sure, absolutely.

Christian (11:07.168)
And so people go to undetectable AI and we have several different solutions. One of our enterprise solutions is authorship matching. And what that essentially does is I can go and put some pieces that you've put together, Michael, and go, you know, I love the way Michael writes, right? And I want anything that I generate with AI to look like it was written by yourself. so that's one of the things that we're working with.

in terms of enterprise partnerships is allowing people to essentially write in their voice as well as not get hit by that penalty of this is AI generated content, which at our core, that's really what we focus on is making sure that people are able to use AI without negative repercussions and essentially arm them with it as a tool rather than potentially opening themselves up to a liability by using AI.

Michael Bernzweig (11:56.629)
Thank

Michael Bernzweig (12:01.328)
Yeah, and I think, you know, if you look at search engines like perplexity, for example, I mean, you know, I think from the founding, think they've always aimed at more of like an academic angle on AI in terms of, you know, footnoting and documenting all of the resources and things like that, that they're pointing people towards. And I think, you know, to that end, I think there's a lot of...

a lot of value in clarity, but I think on the flip side of it, we're seeing so many new uses for AI and I think people's heads are spinning in a lot of ways. think, you know, while clearly AI automates a lot of things and it's a fantastic tool. mean, anyone that's doing

programming or development, using some of these new technologies in their toolbox is gonna help them get to where they want to get to faster. But I'm not so sure that the fear of being replaced by AI is a valid fear, because I think a lot of these different technologies are assisting people and maybe helping to make higher level

opportunities available but I don't know that it's replacing. What are your thoughts on that do you see?

Christian (13:29.72)
Yeah, you know, I think six months ago I would have agreed with you a little bit more than I do now. You absolutely are going to need a captain at least, but even that's kind of being challenged by some of the new agentic AI. I don't know if you've heard that term, but essentially where, you know, all of these AIs work together to accomplish a task and...

Michael Bernzweig (13:34.842)
Yeah.

Christian (13:52.812)
go, you know, scrape the internet and basically figure it out, have free reign to go and figure it out. And there's that's an oversimplification a little bit, but, right. And so, yeah, they are. I just used a, going back to programmers, for instance, I know you had brought up this as a tool to help them. I honestly believe that's going to be the tool that keeps them relevant. If you are one of the people that that's

Michael Bernzweig (13:52.964)
Yeah.

Michael Bernzweig (14:00.015)
Yeah, I mean, AI agents are all the rage. And they're real.

Christian (14:21.506)
does not learn how to use one of these AI coding tools, I think you're gonna really struggle going forward. And we're already seeing some of the implications of that specifically for programmers and developers. And what do I mean by that? One of the tools that we have coming out right now is an AI auto job applier. So essentially, goes applies to a bunch of different jobs. You just hit play, you set the criteria of what you're relevant for.

goes and creates custom resumes, custom cover letters for each job that's being applied for. And what we, yeah, it's, we're really excited about this product and we've had already some kind of inbound interest for that product from educational institutions. And they're specifically telling us right now that, Hey, we are having trouble getting our programming or kind of

Michael Bernzweig (14:56.295)
That's a huge space, yeah.

Christian (15:15.766)
Right out fresh out of school CS majors place right now because of a lot of downsizing that's happening right now. There's, there's a little bit more to that. Right. It's that's one of those where one of those professions where junior programmers really do have a much lower output than maybe a senior programmer would. Right. And so that highlights to me and I'm.

Michael Bernzweig (15:35.43)
Yeah.

you

Christian (15:41.526)
Obviously kind of taking that and running with it a little bit that highlights to me that we are already starting to see that shift of not necessarily all programmers are going to be replaced, but hey, we don't need 50 programmers. We need 10 who are really good at using an AI tool. I just went and used one of these AI code writing softwares when, which I'm not technical. My background is marketing, finance and

kind of strategy, um, being the CEO of the company. And so I don't know how to program myself. I understand the logic and how, you know, AI works and how computer programs work. I went and said, Hey, build me a page that has this, keep all of this in mind, make sure the user journey looks like this. And it gave me a usable version of that. Like just it's.

Michael Bernzweig (16:09.969)
Right.

Michael Bernzweig (16:32.871)
Yeah, I'm right there with you. I mean, I've done the same thing using tools like Replet and some other software solutions. you know, a guy like me, you know, with an idea or an entrepreneur with an idea can use an AI coding tool and get to an MVP just like that. Wow. Now, I think I think what it what it says to me is that the

the game field is changing and it's not gonna be far off before code is no longer a moat for a company. So if that is the case, I think branding and everything else that goes around it becomes even more important for software companies to really have their hands around and to be sure that they...

have differentiation other than the tech stack or the code or the product behind the scenes because at the end of the day that's what's going to help differentiate them from their competition.

Christian (17:39.18)
I completely agree with that assessment. Even with the launch of DeepSeek, we saw a fundamental shift in the way people were looking at AI companies. And it was almost overnight, because we do chat with a lot of VCs, and I have some friends in that space as well. But it really went from the word ChatGPT wrapper, or if you were relying on ChatGPT to build your product as a not defensible, no moat.

And the only valuable kind of entity within that equation was open AI, creating chat GBT. And then as soon as deep seat came out, it really shook up that whole narrative of, that these foundational models are super proprietary with a super high moat and impossible to replicate. Cause deep seat did so and at a very low cost and very quickly.

chatting with some of my VC friends, going, we're actually looking at some of these wrappers again, because our hypothesis has changed a bit towards what's the best user experience with brand loyalty, and we believe that the base LLMs are now commoditized.

Michael Bernzweig (18:51.527)
Yeah, and if you look at the processing power required for some of these different platforms, that's a game changer too, because if it costs way more to serve up an AI search than a traditional search, it makes it prohibitive to scale. But on the flip side, if you can get your hands around the cost, almost like...

you know, if you were to look at traditional Bitcoin versus Ethereum and other tech like that, you you really have an opportunity to get something to market that is not only generating the results that the customers want, but on the flip side of it, at a cost effective price to scale and earn margin on.

Because it's not like in the in the 90s during the dot-com boom when you know you could spend all kinds of money and not earn profit, know now profit is is You're seeing a little bit of that, you know people are trying to trying to gain market share But I think the bottom line is is the bottom line and I think you know all of these different AI companies are trying to figure out

Christian (19:47.534)
think we're seeing a little bit of that right now, yeah, for sure.

Michael Bernzweig (20:04.614)
how do we monetize what we're doing? I I think you've seen perplexity playing around with some ad formats. I think you're starting to see some of the AI search play around with some formats, but it doesn't matter, know, monetizing it if your cost of providing the service is not in line, so.

Christian (20:25.902)
Absolutely, and that goes same with R &D to get to that point, right? If you're spending billions of dollars on developing one of these products and then at the end of the day the total addressable market isn't quite what you spent on it. Sorry about that. Then yeah, you're in trouble.

Michael Bernzweig (20:47.505)
Yeah, so from where you sit, which is a very unique vantage point, and obviously you've been there and done that a few times. This is my second startup. I sold my last company a couple of years ago. What trends are you seeing of how people are using AI-generated content across

different industries because obviously your platform is used by users in many different countries, many different industries, many different markets.

Christian (21:20.098)
That is a tougher question to answer than I would have anticipated. But the biggest one that I see where it's really making an impact you can put in the dollar terms would probably be within programming and coding. We've adopted that internally. We're starting to use these AI programming tools, which have their limitations with complex code base.

Michael Bernzweig (21:33.776)
Okay.

Christian (21:42.398)
I guess that's kind of another piece of information that's super important. If you have a very complex code base and you have kind of strict guidelines that all of your developers follow, the AI does struggle, or at least our experience has been struggling with getting that AI to adopt the way that you do your programming as well as...

consider all of the different dependencies that you have already built out on your website through your complex kind of setup. We are seeing a lot of blog post and kind of search engine optimization be supplemented with AI. So content generation that way.

Translation has been another interesting one that has really been getting a lot of uptake with AIs using AI translators rather than traditional natural language processing translators. It's all over the place, right? I I wish I had a better answer, but I could probably spout off 40 or 50 different things.

Michael Bernzweig (22:44.07)
Sure.

Michael Bernzweig (22:48.497)
Yeah, and obviously you're gonna start to see certain areas kind of bubble up in terms of what the different value is provided by a platform, but at the end of the day.

It's obviously important to see where it's adding the most value. and who is using the product? Is it mostly SMBs? Is it enterprise level clients? Is it marketing people? Who's using the solution?

Christian (23:21.174)
Yeah. So content creators are our primary demographic. So it's somebody that's going and putting out search engine optimized content. It's a lot of consultants use our product who are serving or at least have AI already within their workflow. And a lot of people reach out to us and go, you know, Hey, we were using AI or some form of Grammarly even as an AI tool, right? To help kind of expedite our

spell checking or even our writing process. And now when we start serving something to clients where we use Grammarly, it flags as AI generated and we get, know, our clients are in an uproar going, well, I just paid you $5,000 to get this done. And you used AI assuming that because AI was used in the process that it's not a valuable service that was provided. Right. And so it's essentially being able to protect that company's work.

their ability to serve an AI product without those negative ramifications. We have SaaS platforms as well, email providers, et cetera, that are coming to us and kind of email SaaS companies, bulk email SaaS companies that want to humanize their AI generated content that they're serving to their end user. And that would probably be the primary use cases.

Michael Bernzweig (24:43.928)
And what about educational institutions? Do you see a market there or not so much?

Christian (24:49.558)
So because of the humanization aspect of our business, we've had trouble kind of breaking into the detection space, right? I think it's a little, we've had trouble kind of showing our credibility by saying, hey, we have something that's able to AI generated text. Well, on this side of the table, we're saying, hey, we have something that can humanize the text and make it not detectable.

Michael Bernzweig (25:18.735)
Yeah.

Christian (25:18.988)
Right? And so that's been a little bit more difficult.

Michael Bernzweig (25:21.722)
And I would think academia is really looking for something black and white, right? There's not a whole lot of gray.

Christian (25:26.83)
Correct, absolutely. Yeah, and it's gonna be interesting. I don't know that AI detection is the way to go necessarily for academia and there's so much that's gonna be shaken up by these.

AI providers. mean, I just read something, a research report where somebody had created an AI agent that goes in and it took an entire course, online course, and completed it with no input from a human. Right? So somebody, they essentially signed up for this course at a very reputable school, put the AI agent, basically had it go through and complete the course. And this person, you know, never logged in to

directed or learned any of the coursework, right? It was just completed completely by AI. I see pop up on my Instagram providers that will essentially have a little button that you can click at the top right corner of your Chrome tab and it goes and it answers every single question on a quiz that you're taking for your university, right? And it gets 90 to 100 % on that exam. And so I think we're gonna see a lot

of shake up there, if not a complete redirection, which it's such a big boat, right? It's hard to turn. Education's been the same way for so long. But they're going to have to, or the value proposition's just not going to be there.

Michael Bernzweig (26:51.152)
Yeah, brings up so many questions, right?

Michael Bernzweig (26:56.399)
I feel bad. I feel like we're opening up more questions for the audience than answers.

Christian (27:04.072)
It's an interesting time and I'm sure I will look back at some of my predictions six months from now or what I'm talking about six months from now and some of them will be right and some of them will be way off but it's just changing so rapidly in this space it really is hard to predict.

Michael Bernzweig (27:17.5)
Yeah, and it's an exciting time to be alive. mean, you know, we're seeing so many amazing things that, you know, I look back at the businesses that I've run over the years and just the volume of people and processes and structure and things that were just done manually that today, you know, we're just in automation. You know, it's crazy.

Christian (27:21.272)
Absolutely.

Christian (27:44.75)
The disruption is very real.

Michael Bernzweig (27:48.069)
Yeah, absolutely. So from where you sit, which is in a really unique vantage point, if you were to bullet point a few high level predictions as to what you see coming, what do you see coming at us?

Christian (28:07.746)
I think we're going to continue to see, I think what we're really gonna see take off is more niche down and why Combinator talks about this as well and has put out a piece about this, but vertical AI agents, which it's essentially niching down and one by one trying to replicate an entire job role or an entire.

Yeah, all of the processes within one specific job role, whether that be law, whether that be accounting, whether that be programming, but maybe even not even just programming front end programming. I think it's going to be quite a while before we see somebody going and saying, being able to go and say, Hey, do all the marketing for my company and tell an AI agent to do it and getting any meaningful kind of results. I do think once we start.

focusing down on these individual problems that need to be solved and then solving them one by one, you're gonna have several winners, not just one winner like ChadGBT, you're gonna have dozens, hundreds, if not even thousands of these winners that are very specifically creating these AI agents to optimize and solve for a very specific repetitive task.

Michael Bernzweig (29:24.206)
Interesting. And as far as undetectable AI, if you were to summarize in a sentence or so, the problem that you're solving for your users, what would that be?

Christian (29:36.312)
We are allowing users to take AI.

and really level out the playing field with some of the larger, what larger institutions have been able to do with tons of people. That's our goal, and we are even working on some of these very focused AI agents coming out in the near future here.

Michael Bernzweig (30:00.453)
I love it, I love it. And as far as the trajectory of Undetectable as an organization, where are you guys at in terms of, take it you bootstrapped it originally and then you went to venture or you guys, where are you at in terms of funding?

Christian (30:16.482)
Yeah, absolutely. So my last company, Chatterquan, I did go and raise some venture capital before that was acquired. But with Undetectable AI, we're completely bootstrapped up to 55 employees and continuing to grow there and just hit 15 million users in about two years now.

Michael Bernzweig (30:37.018)
I love it, I love it. And do you envision some next steps in terms of taking on funding or do you think you're pretty set in terms of organic growth, product-led growth and all of that?

Christian (30:51.54)
We've been very ROI focused since the very beginning. And so we are trying very hard not to put ourselves in a position where we would need to take on funding, but if the right partner were to come along and we thought they could help take us to the next level, we would consider taking funding, but it's not in our immediate roadmap.

Michael Bernzweig (31:10.864)
Yeah, and I think what's so interesting, and I've seen this with so many amazing ideas, at the end of the day, when you have a recurring revenue model of any sort and you focus on ROI, you can do things that are magic in SaaS that you can't do in other industries, product-based industries.

And I think just to that point, just what you said there, think finding the right partner that can add some other value besides dollars and cents to help get what you're doing to that next level is really the value that an organization needs. And I think that makes a lot of sense, so very smart.

Really appreciate the deep dive here. think we have covered a lot of great topics and I think we had a bunch of questions come in from the audience before the podcast and I think we touched on a lot of different questions where we saw many, many of the same questions come in from the audience. So we'll obviously leave a link in the show notes for anybody that would like to check out what you and your team are doing there.

and I really appreciate you joining us on the Software Spotlight today.

Christian (32:31.564)
It was a pleasure, thanks for having me on.

Michael Bernzweig (32:33.831)
Yeah, thank you.