SaaS Backwards - Reverse Engineering SaaS Success

Ep. 174 - How This AI SaaS Landed 70 Customers—Fast

• Ken Lempit • Season 4 • Episode 27

Share episode

Guest: Adir Ben-Yehuda, CEO at Autonomy AI

The go-to-market playbook for AI SaaS is being rewritten in real-time, and those who cling to old models risk being left behind.

In this episode, Adir Ben-Yehuda, founder and CEO of Autonomy AI, joins host Ken Lempit to share how his team went from zero to 70 customers in months by ditching outdated frameworks and building a brand buyers can trust.

We unpack:
✅ Why “Crossing the Chasm” no longer applies in the age of ChatGPT
✅ How brand marketing now beats lead gen in AI go-to-market
✅ What it really takes to convert skeptical enterprise buyers
✅ The shift from SEO to “share of response” in AI search platforms
✅ Why every SaaS GTM leader must become an orchestrator—not just a doer

Adir also breaks down the sales motion that helped Autonomy scale so quickly—and shares how his team leverages live demos and social proof to close deals in a single call.

If you’re a B2B SaaS CMO or CRO navigating AI adoption, rethinking pipeline strategy, or looking for a more effective way to win technical buyers, this episode is your cheat code.

---

Not Getting Enough Demos?

Your messaging could be turning buyers away before you even get a chance to pitch.

🔗 Get a Free Messaging & Conversion Review

We’ll analyze your website and content through the eyes of your buyers to uncover what’s stopping them from booking a demo. Then, we’ll give you a personalized report with practical recommendations to help you turn more visitors into sales conversations.

And the best part?

💡 It’s completely free.

No commitments, no pressure—just actionable advice to help you book more demos.

Your next demo is just a click away—claim your free review now.

Welcome to SaaS backwards, a podcast that helps SaaS CEOs and go-to market leaders, accelerate revenue and enhance profitability. Our guest today is Adir Ben Yehuda, CEO at Autonomy AI, an AI platform that provides a fleet of context aware agents to take on front end development. It automates repetitive coding tasks and reduces tech debt to accelerate product releases serving high growth startups and enterprise RnD teams alike.

Ken: Adir, welcome to the podcast. 

Adir: Thank you. It's great to be here. 

Ken: Yeah, really excited to record your episode. But before we get started, could you please tell us a little bit about Autonomy AI, and your background before joining the firm? 

Adir: Sure. So as you mentioned, the CEO and Founder of Autonomy AI. My background is coming from go to market actually, so not in that

code space, but mainly around selling to sales organizations and marketing organizations. I've took three companies from zero to over $10 million in revenue. For some reason I like to join very early on, right? It's like, yeah, oh, you have no product, that's great. Let's make something out of it. So I'm very much enjoying this, and we did the same thing,

I thought to myself, let's found a company. So we started Autonomy AI. We in the company, we have two other co-founders,

my CTO, Tammuz Dubnov and Eric Feingold is our chairman. We're as, as you had mentioned, we're focusing in the code side of things and helping companies actually accelerate their ability to produce software and produce.

So, been around since September, 2023, launched the product in April, 2025 and scaling very quickly. 

Ken: So that's pretty quick founding to launch. Are you eating your own dog food here? 

Adir: Yeah, of course. That's, that's a key part. So it's the paradigm. I think what a lot of people when they look at agents maybe kind of missing or don't really realize is there are two things, right?

A, we obviously build our own product, right, with their own product. But the second thing, agents are a living organism, if you will. They learn and adapt all the time from every run they do, and kind of build themselves. And that's a key part of what we had, right? So we kind of build a foundational that the agent will work on. Once we start building our product.

The agent tech became better and better and was able to produce way better results in month three than month one. 

Ken: My very early career I was a coder and project leader on big software projects for IBM and then Citibank. So I've been decades long student of the development world.

So it's exciting to see some of the promises that were made to us 40 years ago come true. So, 

Adir: yeah, 

Ken: this is, this one's a particular interest for me. let's talk about go to market and, when we talked in our prep I brought up the book Crossing the Chasm, which we're both fans of, but you said, you know, the, the model might be outdated for today's go to market.

I think we need to dig in a little bit about what's driving companies at different maturity levels to adopt AI solutions like Autonomy AI. 

Adir: Yeah, so maybe, maybe let's talk a little bit about the concept of crossing the chasm, right? It's kind of divided in two buckets, two main buckets, like the early market, and then kinda the mainstream bucket.

The early bucket you have like the innovators, the early adopters, they're still there, but then the mainstream bucket, you kind of have the early majority, the late majority, and then onwards. And when you talk about the personas of the early majority, the late majority, they're usually their profiler.

People that are pragmatic, pragmatic buyers, they don't want to hear about your roadmap. They don't want to hear that. One day. You'll be able to also do that and just enjoy the ride and, and join the company, like buy your product. And then you have the, the skepticals, right? Like people are like, it's not gonna work.

Or I'm too conservative for this and I'm like, I'm too worried. I don't know what this thing is gonna do. And if you look at what happened, I think that the pivotal moment that happened was back in 2023 when OpenAI kind of launched ChatGPT and it became overnight the most adopted application on Earth, right?

Like in a few short days, this entire model kind of collapsed. Because when you look at the second bucket and like the late adopters, skeptical buyers cannot be skeptical. Everyone is using it, right? If you're using it in your own household, you utilize it for your own things, you kind of look at it for work and vice versa.

And if you look at the pragmatic buyers, I mean.

What reason they have not to be pragmatic. They look at what AI can generate and it kind of removes the process from getting the result and gives you the result. So it's very easy to see the how, like you can be very pragmatic and how efficient it could be for you.

What I'm looking at right now from Autonomy's AI point of view is like everything is collapsed, we're getting every, everyone from everyone. We're getting like mature companies. We're getting like big enterprise, we're getting small organizations and everyone is kind of jumping on everything and this entire paradigm of like, okay, we need to be very slow.

We have innovators, we have people that we want to tackle. It's in my opinion, I see it kind of collapsing and falling down. 

Ken: I, I think that there's there's definitely some like fear of missing out here, right? That you could be eclipsed even as a large scale enterprise if you're not adopting some of these tools, right?

That your, your ability to compete, bring new product to market could be dramatically impacted if you are not using these productivity tools. Does it mean that we have to do different marketing to these like enterprise more conservative firms? Or do they come to us, you know, out here on the forefront in AI?

Are, are these folks looking for these solutions or do we have to go instigate the conversation? 

Adir: They're a hundred percent looking for them. I think one, one of the things that we learned and we saw is you just need to be out there, right? I think that there's kind of also a different paradigm in, in the way we look at marketing.

Instead of like reaching out, doing a lot of, you just need to be with billboards and with the headlines, and people are looking for it constantly. And the reason they're looking for it is because they're getting a lot of pressure, even in big organization from management. From board, right? If you go, and probably a lot of people from your listeners are listening and hearing this and saying, Hey, it's actually happened to me.

You sit in an executive meeting or you talk with your boss and people are saying, what are we doing with AI? And you got to have a good answer. Or what other AI tools out there can help us more efficient or increase like the cost of acquisition if it's in marketing, or essentially increase velocity of development or reduce bugs in the development cycle.

We've all heard this in the past year in some way, form or shape, and everyone is actively looking. It means you just need to, from a marketing perspective, you just need to be more out there and actually, instead of thinking like a sniper. You need to think more broad, right? And make sure that more eyeballs can see.

Ken: So this is really the case for brand marketing right now, right? We don't have to do demand or lead generation. It's really brand, brand development. Create reputation in the space you want to operate in. 

Adir: Exactly. While crossing the chasm is kind of changing, the, the psychology of some individuals will still be there.

They will still like to be, they will buy earlier, but they will still buy to want to buy from a brand who's trusted and has good social proof. So there are two things here, right? Like, let's not be fooled. People will buy earlier, will adopt things earlier, but the same people who are late adopters, skepticals or maybe conservative, they will buy earlier.

But they would like to buy from a trusted brand. That's why you need to invest more in like the brand marketing and social proof around. 

Ken: I think that's a great takeaway for go-to market teams and for at firms like yours, right? It's, it's not that you don't need to do marketing, but you need to do the brand work.

Establish that foundation, let the demand come in as it will. Because there will come a time soon. Is it a year or two years out that we're gonna be back to creating awareness and demand for the solutions as well?

Let's kind of move on a little bit. I actually wrote my first LinkedIn newsletter on this topic today.

Yeah. Congrats. Thank you. Yeah. Well, we're, we're out there with a lot of content marketing, but I finally decided to take up the bid on a LinkedIn newsletter. this is a really important topic. So AI is potentially reducing the need for people in the workforce, right? And I guess my question is how do you see the role of knowledge workers evolving?

And, and when we talked about it, it was like from doers to orchestrators of, of AI driven process. But I, I also wonder if there's something else at work here, which is kind of like, the analogy I use today was. You know, everybody's seen a new highway built or a highway, they make more lanes, more traffic lanes on the highway, and the traffic always seems to take up the available highway capacity.

And I'm wondering, do you see actually a reduction in workforce needed, or is the role of work gonna change? And are we gonna need more orchestrators than doers?

Adir: Well, first of all, it's a very, you know. Loaded emotional question, I would say. Right? It's a topic. Yeah. Well, welcome to a podcast. 

Ken: Welcome to the podcast.

Adir: I think it's, it's a fair question and I think it's a good question. I talk with a lot of people and I'm in front of many, like. I'm doing a lot of like conversation and lectures about this space, and it always ended up with, with this, right? Like, what are we gonna do next? And, and, and I think the way I see it and the way I believe like the next 10 years are gonna unfold, I think.

Yes. More orchestrators than doers. You, first of all, you, you would be expected from the organization you work for and you work in and you're part of, you're, you're gonna be expected to generate more and to produce more. So by definition, like you'll need to find all those hacks, right? To adopt all those hacks.

And those hacks are gonna be mainstream. And the second thing is, instead of focusing on processes, instead of looking at like, what process can I do to get to an end result, you have to be thinking about the end result. So kind of orchestrating a bunch of AI tools and bunch of AI solutions from that aspect, right, that you would be able to control and get you the result itself.

It means you'll need to be more creative as a, as a human being

as a person. It means you need to be more crafty. Think about like the next steps and be more strategic on how you can all orchestrate. Same as managing employees, it's same thing, right? I think the second thing, and that's another part then that I'm looking at to just open up the market to more entrepreneurship.

The ability to actually generate more companies, reducing the barrier of building a company and building a solution. I just stuck with a friend of mine and he is like, he took. True story from this morning. He went and took his haircut today, went to the barber, and the barber is like, like a small shop downtown.

New York, downtown Manhattan. And he talks with the barber, like, so, like the barber asked him, sorry. He's like, so what are you doing? He's like, oh, I'm in an AI. He's like, oh, I have this problem. I'm trying to get like, appointments and I can't find the right solution, the right application. He's like, don't worry about it.

He sat there in the store, in the shop. Two hours, build them a nice application with like, where, where it can book things. They can book like different different like, appointments and things like this and track them and track payments and to do the rest of those things. So that's a key part of like what AI can do for you.

So opening, moving people from doers to orchestrators for opening also the market to the ability of more entrepreneurs and more solutions out there as well.

Ken: This is sort of an emotionally loaded topic, right? The continuing relevancy of knowledge workers in the workplace. And I, I'm really an optimist. I think that we, we need to land on the optimist side of things as, as you just tried to do.

It's gonna create more work, right? There'll be more demand. For more tech to be built to solve more business problems. And you know, Mark Benioff's statement aside that he is not hiring any engineers at Salesforce. I'd like to take a look at that about six months from now, it'll be a year. Almost since he said that.

And my bet is there'll be more engineers on staff than there were six months ago. But they'll be doing different things and the work will in some ways be better. You know, if you take some of the rote work out of systems development. Imagine how much more creative people could be. So, I think this is actually a huge opportunity and sure there's gonna be some people who don't have the job they have today, but I think there'll be more jobs in the end.

And I don't think there's been a technology development that's reduced growth in employment. I don't think we have that yet. 

Adir: I think there, there is. Gonna be a dip for the short term. And I think for the long term, there's gonna be like an evolution. If you look at like, I think the, the next, the less big thing that happened in this equivalent, which is the Industrial Revolution, right?

What happened is like, yes, the workforce changed. People in the short term lost their jobs, but then the then became a boom of like different occupations, right? And if you think about it, what happened right after the industrial revolution, the entire concept of psychoanalysis started Freud, right?

All of a sudden people started saying, oh, I have a little bit more time. I can think about myself. I can think about my feelings. So I think we'll see the same evolution here. The only thing, the one caveat, I would say this is gonna be way quicker. Your comment about Mark Benioff, about Salesforce? I agree.

Probably they're gonna have more people, but the technology is not quite there. But in the span of two years, I would bet you'll see a reduction in force. Everything is moving in a different pace. 

Ken: Yeah. And adoption pace and impact gonna be much greater than anything we've seen before. Let's talk about your launch and go to market. So, you guys went from no clients to 70 companies using the product in just a few months with significant ARR. So let's talk about the go to market for Autonomy AI. What did you do that's enabled such rapid adoption? 

Adir: So let's, let's divide it into two parts, right?

Maybe talk a little bit about our model. We're currently a high-touch sales motion, meaning people come to our website, they book demos. They book meetings, and we like take them hand, hand in hand, right? And we divided it into like, let's divide it to two parts, the top of the funnel and how we deal with the deals that are coming in.

From the top of the funnel perspective, when we did the initial launch, we essentially built a lot of content around it. It wasn't talking about us, right? It was talking alluding to us on our website, but we built content that people actually cared about and was timely for them. And the way we did is we basically asked our prospects, like our future prospects in our friend groups and friend cycles. What did they talk about internally? What did they talk about in management meetings? What did they talk about with their teams? And we built a set set of content around this, right? That was the initial launch. Second thing we did, and I recommend this to a lot of people here, is Product Hunt was super successful, not expensive.

Got us a lot of lead, a lot of exposure, and got us a great social proof. 'cause we kind of scored in like in the top four of companies there, and those are the key things we did. Then the second part, we enable people to come to the website and book meetings very quickly. The first thing they had to do is put an email.

I capture the email , and then I can let them go down the funnel of like booking the time. So that's kind of like something that helped a lot. The second part, and that's a key thing up to now when we have a meeting and that, that that part comes to the conversion. It's very easy on a 30 minutes call to see if it's a go or no go for the prospect.

We build the product and we kind of try to orchestrate the demo part that the prospect could see within 30 minutes. If this is something that is valuable for them. If this is something that could generate a linear ROI, meaning I use it today, tomorrow I get the result and I don't need to go through a hoop to like understand how we created an impact from that matter.

And I think the third part, we can install this right now. So Ken, I would go on a call with you. I would say, great, 15 minutes in you like this. Let's open your computer. Let's write on, it's like literal three minutes. Let's check it out. Right? And kind of helped us. Going back to the initial part of our call, all the pragmatic buyers saw it very instantly 

We kind of built our rapour and our social proof on the call itself and didn't hide behind the scenes. So that was a key part of our, of our initial success. 

Ken: Is there any place in these organizations you see this getting the most use, like any functional area of your client organizations? Are there SaaS firms building their apps? Are they enterprises building internal, internal tools? I mean, where is this getting the most traction? 

Adir: So today the product focuses on frontend and we are essentially mimicking and building an autonomous frontend developer for organizations.

And we see it like from a company's perspective. We see a lot of cybersecurity. Data companies that are coming to us, right? Marketing tech companies that are reaching out to us and that the people are utilizing them are usually CTOs. Or have CPOs, Chief Product Officers that wanna accelerate the the flow.

We even have a lot of CEOs that in, in larger companies that are involved and part of the buying process and the adoption process internally. 'cause the way they think about it, this is like a mandate that I have, I need to move faster and like in a more lean way, this is something that could help. So we see it across the spectrum.

And the main place they want to utilize is when they want to accelerate deliverables on software and they wanna say, oh, we have a deadline. And we want to get like, we wanna move faster before the end of the quarter or before a competition, or we wanna increase revenue, so we need to release this feature.

That's where they adopt autonomy very quickly and utilize this. 

Ken: Makes a lot of sense. Kind of shifting gears maybe a little bit here. And this was surprising to me, but you mentioned that fundraising was a bit of a challenge early on, and yet, you know, you kind of started out, AI was, you know, the rocket motors were lit. So, what were the objections that investors had and how did you ultimately prevail and get, get the funding?

Adir: So like everything, right? I think AI was the key thing, but like a lot of VC, a lot of investors we talk with. Did not really know what AI is. They understood the buzzword, they understood what LLM is. Right? Maybe they realized, and I'm talking back at September, 2023, maybe they realized their agents starting to like be out there, but it was very early on, I think we're one.

I was just talking with a big consulting firm with their CEO. I'm not gonna mention the name. And they had mentioned that agents have been around for eight months and I'm like no. We had an active agent in January, 2024, which is like a year and a half ago. So one of the key things is like, think about it, we're come, we're coming to all those VCs.

Yes, we have a background. We build a few companies. We were able to scale a few companies, but we're coming with something they've never heard about before. And with this big promise of like. Guess what we can do, what a human being is doing just better in scalable way right? And it was like, I'm thinking about it back then.

I'm like, where did we have the nerve, nerve even like to come, come to them, right? And we're like, we're really, it's like, what are those agents like? We heard about them, but is it real? Is like five years in the future? We were not sure. Probably someone who can build them. It's probably not gonna be you and you probably need a billion dollar to build those kind of things, right? Fast forward to July 2025. Agents are like the main technology. Everyone are utilizing them. But at the beginning we had to kind of go in and say this is something that could actually work and educate our investors on what could actually work. A good hack that actually helped us, and maybe it could help your listeners as well.

I went to some of the investors portfolio. To their website and kind of targeted companies that I think could be early adopters for us, and took the LinkedIn names of their CTOs and told them, if you, if you want to trust us, talk with this person, I'm happy to hop on a call with him. Let's be proactive about it.

And we kind of started building the social proof. We kind of start building the confidence on what the product can do, even from a conceptual perspective of like, oh, this is actually makes sense. It's not complete BS, right? So that's something we were able to kind of tackle head on. But really the main challenge was, how can you explain to someone what's gonna happen two years into the future, which is now, right? 

Ken: And it's interesting 'cause VCs have this reputation for being so forward thinking, and I think it's actually the entrepreneurs have to drag them into the future. I love that hack of going into the portfolios and trying to get a vote of confidence from some of the CTOs. That's a great idea.

Adir: it actually works. It's great. 

Ken: have you actually had any traction with investors, portfolio companies? 

Adir: Yeah, quite, quite a few. I think it's a great way to also build your pipeline and convert and convert deals. So yes, and, and even later on, later stages when we kind of like talk with different VCs, they said, Hey, part of the DD is for us to connect you with the couple of portfolio companies.

I'm like, bring them on. It's like if this investment is not gonna pan out. At least I'm gonna have two new clients, or at least two new leads or prospect. So yes, it's working again, goes back to the place that everyone is, everyone wants to try everything right now and especially when their VC is talking with them

right? And I'm saying, I have a product for you that can help you become more efficient and scale faster. Like you're not gonna say no, you're gonna say yes, I'm gonna check this out, right? So that was very helpful.

Ken: We, we did touch on this a bit, but I, I want to go back to what's creating inbound interest.

You know, you're not spending a lot on marketing at this time. Where are you placing your bets that either organic or paid at this point? You know, where, where are you putting yourself out? You mentioned product hunt. Where else are you active in go to market? 

Adir: The challenge, as we had mentioned and defined is like, we wanna get a bigger chunk of the market. We need to build rapport, we need to build social proof, right? People still want to like, buy a product that they know. We're doing a lot, a lot of content, that's the first thing. And we're trying to build content, it's timely, so we kind of build a mechanism.

Everything is buzzing right now, within a 24 hours using AI, of course, we'll generate pieces of content that are timely there are very curated, and we see a huge amount of inbound traffick against them. But it also helps us become like basically, thought leaders and specific topics. The second thing, we're collaborating with a lot of companies and brands in our space, bigger brands, right?

They wanna build the content. They want to become thought leaders. They want to bring things that are innovative. So we're doing a lot of cross content collaborations. We're doing a lot of podcasts. Dinners and events so people could see us and see, this is the real thing. I think it's big in this new AI era, right?

Like people are, people are raving, they wanna move into like the social experiences and be with one another. It's a huge thing. And the third part is we're focusing, we kind of removed SEO almost completely. We were focusing heavily and on agentic or basically AI engines optimization, right?

Focusing on them a lot because that's where our audience is. They're searching things not on Google. They're looking at ChatGPT or Perplexity or Claud. And it's a key, key driver of traffic for us as well. 

Ken: I couldn't agree with you more. I mean, even before the rise of search on chat, SEO was almost a dead heart.

And I think now, it always kind of amazes me that there's still people trying to sell traditional SEO services. But how are you optimizing for share of response? I don't even know what you call it. Share of response on a chat. 

Adir: So the key part is to figure out the niche questions that that's what worked for us, right?

So if I will ask, how can I help, or if I would try to kind of compete, quote unquote, on an answer of how can I optimize my co-generation in my organization? It's gonna be very hard for me, but if I'm gonna ask questions of like, Hey, I've fronted components in React, which people are asking, you'd be surprised and I wanna optimize them, that's where I can shine.

Not a lot of companies are competing for this, but a lot of people are asking about those kind of things, and I'm unique and add unique content around those kind of things, right? The new equivalent of rising above the noise is actually. Being, below the noise and, and understand what people are, are searching for

that is niche and very focused on your aspect and your spectrum. 

Ken: Where are you getting those insights? Into the questions that people are asking? 

Adir: We use a third party tool and we utilize this tool. It's, it's been great for us and it's really working, 

Ken: Fabulous. Let's land on

a case study kind of conversation about Autonomy AI. You know, you've defined it pretty well. You're transforming front end development saying that days of work could become minutes. Can you walk us through a specific example? You don't have to mention the client name, that's fine. But how, how is Autonomy AI

actually reducing or accelerating the software development process? 

Adir: Yeah, so I, I have a, I have a bunch in, in this specific use case, a lot of companies, they have their backend ready, right? They have the engine, but they want to kind of move to the new AI UI, right? The AI user interface where people are getting used to.

So a more chatty experience, more back and forth, more dynamic experience. And a lot of our clients. Are coming in and saying, Hey, we've planned this entire redesign of our platform and it's gonna take us like three months to, to deliver on this. 'cause we need to generate the designs. We need to generate the code.

There's gonna be bugs and there's a whole iteration. And a lot of them are coming to us and saying, Hey, we have all the designs for you. Can you help us accelerate everything? And specific one we just did where we just helped recently. They planned for three and a half months. We finished everything within a week.

They basically did an entire redesign of their software to be more fitted to the AI era, I would say, right? For their new clients, for their current users as well. That's something we kind of help. But guess what? The result was also bulletproof, meaning there was no QA needed. 'cause an agent is doing this and everything is more accurate and more to the point.

And from now on they can actually build their platform and it's more AI ready in terms of the code. So every time they want to use us or any other solution in, in the space for that matter, the code that is generated would be able to be digested if I will, better in a and more hygienic for AI.

Ken: It's interesting you mentioned they have their designs. Is there an integration with other design tools like a Figma or something else that helps drive the process? 

Adir: It's a great question. So we can do everything from Figma, screenshot or even plain text, right? Even if you have a robust platform or platform that you've been working on for the last couple of years, all you need to do is just write in plain English what you wanna generate.

And we would be able to generate the, the screen, the result without any Figma, without any design contextually to your look and feel. So. 

Ken: So you can refer outside to a existing look and feel. Yeah. And does this also integrate with existing code base? So if I've got some code already in play, can we add add on to our existing code base? 

Adir: Yeah, that's, that's exactly the thing. The first initial run that we do, we build the technologies that can actually understand your current code base and then generate as submission new screen and or as you refer to, update your current code base or update specific components or specific screens that you have.

Ken: Well, that's kind of a brave new world, especially for guys like me that started hand building software so many years ago. If people wanna learn more about your company or reach out to you, how can they do that Adir? 

Adir: So they can connect with me on LinkedIn? I would be happy to connect or go to autonomyai.io register on the website.

We'll, we'll be happy to talk with them. 

Ken: That's great. Well, thank you. If people wanna reach me, I'm on LinkedIn/in/kenlempit My advertising and demand generation agency for B2B Tech and SaaS is Austin Lawrence Group. And if you haven't subscribed to the SaaS backwards podcast, please do so wherever podcasts are distributed.

Adir Ben Yehuda, thank you so much for being on the podcast. 

Adir: Thank you so much. It was great to be here.