Taming Tech Stack Bloat with AI

0:00Even today, when we look at AI, just in general, right? We have a lot of text-to-text, and that's something where we're prompting the AI to give us a solution back or an answer back, which is great. But as we're moving forward and we get more to this text-to-action, that's where it gets really exciting. Welcome to the Talking AI Podcast, where we talk AI with both experts in the field and early adopters. I'm your host, Matt Page, and we're here to demystify AI for you so you can get some value from it. Let's talk some AI. AI is creating a massive blue ocean of opportunity at the application layer.

0:36It's forcing businesses to really rethink how they use software. But in this process, many enterprises are realizing something a bit disturbing, and that's that there's this overwhelming amount of bloat in the tech stack. And this explosion of SaaS tools we've had over the past few decades has led to a lot of inefficiency, wasted spend, countless people having to manage these tools. But lucky for us, we got Vanessa Heppner here, CEO and founder of Altus AI, with experience leading planning and operations at the highest level at Salesforce. And she's seen firsthand how companies adopt and manage technology and where they go wrong.

1:14And now she's leveraging AI to help solve that very problem. But welcome to the show, Vanessa. Thank you so much for having me on. I'm really excited to be here and to be able to chat with you about this topic. It's a really big one. There's a lot of nuance behind it too. And I'm going to start here with an example. I'm actually recording another podcast with this individual later today. And so I'll set it up. It's this guy, Michael Liu, right? So the story goes, someone put out this viral tweet saying how expensive DocuSign is and basically pleading for alternatives just out in the universe trying to find somebody to help.

1:49and Michael being the smart guy that he is, you know, he's an engineer, past engineered meta, Zillow, sold his last company to Stripe. So he's a smart dude. And he just decided, maybe I'll just try to build this with AI. And he did it in two days, right? So he put out this tweet about it and that went viral. And then he's like, heck, I'll just bring it to market. So you have this guy that built this alternative in two days, right? So there's this one side of things where people can now build way more efficiently. I think the other side is like AI brings this whole new novel approach to solve existing problems, jobs to be done, and all of those things.

2:28But I'm curious, do you think enterprises actually understand how like massive of an impact all of this is going to have, especially like the larger enterprises? Yeah, I think they're feeling it, but they're also excited about it too, especially because we're getting so much interest from different customers, potential customers out there who are just excited to become more efficient. And I absolutely saw what you're referring to about building a little lightweight solution really quickly to replace DocuSign. Like I saw that online as well, and I was like, wow, that's incredible. So it shows that there's so much opportunity to start to rethink how we build, how we make things more efficient with our systems, our processes.

3:10And AI is helping us get there. And I always love to talk about like first principles mentality and that type of thinking. We apply that within Altus every day as a team, as we work through different product features, when we're working with our customers. But that first principles mentality is really where AI is taking us as well, where it's, okay, let's rethink things. How do we make things more efficient? And the companies and individuals that are leveraging AI are able to do that very quickly and start to see really great results. If it's like cost reduction or making something more efficient or it's like a very quick, lightweight solution that can solve something in a more efficient way.

3:48And it's really cool to see that. And the folks that are just jumping in and going all in and testing it out are really feeling those results, which is really cool. Yeah. And it's like this, I don't know, it's like this David versus Goliath battle. I feel like we're about to have where you have these very, you know, new entrants, similar to yourself, but you're kind of serving those enterprises, but like you creating an organization today, you are thinking from back to like first principles, but you're integrating AI into your business from day one, right? Versus existing businesses, they kind of got to figure out, oh, this is how my business operates today.

4:27what do I do to kind of work AI into it? And, you know, does it impact my business model, my OPEX, SaaS, or all the SaaS products we have out there? And I think that's going to be the really interesting thing is how do companies adapt? Quick break in the pod. If you're listening to this podcast, chances are you've been thinking about how to actually use AI inside your business. And that's exactly why we built the AI Opportunity Finder. It's a free tool that helps you uncover high impact, tailored AI use cases based on your business, your goals, your pain points, and your industry. No fluff, no generic use cases, just real ideas that fit your business and the ranked by ROI potential.

5:09It takes about three minutes to run and it's like having your own personal AI strategist for free. If you want to try it for free, check out the link in the show notes or go to hatchworks.com backslash AI dash opportunity dash finder. That sounds... I don't know. I'm curious, what do you see companies doing right now on that front, you know, when it comes to AI and everything out there? Like, how are companies starting to adapt, trying to adapt, failing to adapt, I guess is another angle as well. Yeah, adoption is really important. And I think what we're seeing is that the companies that are quick to say, okay, let's either try and test something out internally.

5:47And so we're seeing that a lot where companies will try to test something internally. If it's an AI solution they're trying to build, maybe to add on to their product that they're offering or maybe to try to make things more efficient internally. So companies that are testing that out, that's great. But a lot of times, even in their testing, they end up looking into other firms or companies to help them with that transformation. Because it really is a transformation that they have to undergo. And we hear the word digital transformation, but companies are really leveraging that with AI. And the ones that can do that quickly and kind of test it out.

6:19And then even if they bring in others to help them facilitate that process of transformation, it's really great for their business because, again, ultimately making things more efficient. Or if it's a new product that they're building, a new AI solution that they will be selling to their customers, it's great. But it's really good to see companies just go ahead and quickly start to adopt, try it, test it out. That's why I always tell everybody is just test it out, try it, and I'm sure you're going to love it. Because there's so many great opportunities and ways to rethink things. And again, back to that first principles mentality and just trying to look outside the box.

6:53And AI helps us do that quicker and faster and allows us to be more nimble and agile. And that's something that I absolutely love. And it's really great, again, like bringing back to being more efficient. That's like something to our core as a company. And we just love helping companies be as efficient as possible. And AI helps us do that. It helps us if we take the example of companies having so much, you know, tech bloat, right? Right now they have enterprise companies. We'll use them as an example. We'll subscribe to so many different SaaS assets and tools. And that's just one part of their whole IT portfolio.

7:24If you look at the full landscape, you know, companies that have a hybrid strategy and they've got data centers, there's a lot, right? But how do we lean it out to get more efficient? But again, going back to adoption, I think the companies that are just really quick to just test it out and try it are the ones that have an easier way with adopting it entirely in their organization. It always starts out with testing it out quickly and seeing what opportunities there are in that test and then implementing that across the board. But that's just like what we're noticing is just seeing that the ones that are just quick and open to, okay, let's test this out, really start to bring in that adoption much faster and more precise for what their business needs are or for the solution that they're providing to go to market with.

8:06Yeah. And I think that works at the enterprise level, but also for individuals, like anybody listening right now thinking like, where the hell do I start with AI? I feel like there's 10 new tools coming out a day and you're probably wrong. It's probably a hundred new tools coming out a day. It feels like, but it's the same thing. Like just start playing with the tools. If you can just like incrementally build, try something out. It's amazing how much you can learn in that process. And I think it definitely applies at the enterprise level too. But if we think of, you know, all of these, you mentioned something earlier around companies, not even really knowing the SaaS products that they're subscribed to, or if they're being used and all of this, especially the larger the enterprise, you get these silos in the organization.

8:50I've seen this before where you may have one organization using the same tool as another organization. It's like duplicative in a sense too. And I think that's effectively what Altus is trying to shed some light on is like leveraging AI to identify, okay, what are you actually using? What is being used? But I'm curious, like, how are you thinking about either the product today or like product roadmap wise how you start to leverage ai in a sense to identify what's going on in the org when it comes to all the sas products and licenses they have yeah so when we meet with a lot of companies all most of the things that they always bring up right away is we just want to have that single pane view yeah right like we want to have a pink glass view of their whole like it infrastructure they want to understand what are the things that we're spending all of our money on and we want to understand how it's being utilized within our company so that we can better understand what the ROI is on everything that we're spending for our infrastructure on the back end day to day.

9:52And it's very complex because a lot of organizations have a hybrid cloud strategy. So that makes it very exciting where you have all the SaaS applications, but you also have things like data centers. And so when we're bringing all of that, if we can bring all of that information together, which our AI does, then a company is able to clearly visibly see where they're spending for their, you know, for their technology on the back end, which is exciting. And so being able just to bring that forward for companies is illuminating because they're able to understand how much you're spending and where, see where the redundancies are, right?

10:25But then also when it comes to data centers, how do we make things more efficient, more energy efficient as well? And bringing that, and something that we always talk about is when we're able to help companies be very efficient with their tech stacks and help them reduce costs on things that maybe they're overspending and they don't need to anymore. We're also helping them bring down that energy usage consumption as well, which is something that's going to be a really big topic that I think we'll hear more and more about because technology not only is very expensive, but also the energy usage and consumption there also is very high too.

11:00Our main mission there is to make things as efficient as possible for companies across the board. And our AI is able to do that. And it's really exciting to see it because even today, when we look at AI, just in general, right, we have a lot of text to text. And that's something where we're prompting the AI to give us a solution back or an answer back, which is great. But as we're moving forward and we get more to this text to action, that's where it gets really exciting. And then when we get to the action to action part okay we're just really starting to move into that kind of robotics piece of it but for us we're so excited because as we start to because again right now with ai just that text to text that's great but bringing it more into those actionable insights is something that we're doing at altas and really just trying to push that seal of innovation when it comes to ai because that's where we're moving with all of ai is more actionable so that you're receiving more of those actions right away every single time that you're logging in and see oh wow okay here's an action that I can take and being you being prompted by the AI with an action that you can take or not take right it's something that is we're very excited about and what we see in like the future of AI and how that's moving and again as we look at that trajectory of AI and how it's incorporated in our systems and processes and then you look at things like robotics right so we're really getting more into that actionable really creating more just yeah more action coming out of the AI versus is just I have to prompt it constantly.

12:27So it's exciting to start to see that transition and change across the board. And that's, I think, the big unlock and trend that people are talking about right now because it starts to center around the agentic nature of AI agents, which is like the most hyped up thing in 2025 right now. And there's some people doing it successfully, others failing. But if you think, and I think people struggle to understand what an agent is at its core a lot of the times. And the way I like to explain it is how Google explained it in their white paper. It's like three components. There's your LLM, which is like the brain of the AI.

13:01There's the orchestration layer, which is guiding, okay, what's your purpose? What's your role, your goal, the memory component of it. And then the thing you mentioned is like access to tools so it can take action. And then when you start to build these agents out that have a purpose, can take an action, it's really insane what they can start to do. And the way I always like to describe Altus AI to people is it's like rocket money, but for the enterprise, right? So I don't know if you've thought of this as like my off the wall idea. But if you had the AI agent that was tied to a voice agent that could then go negotiate the contract for you and then, you know, either work it down, cancel it or whatever.

13:41That's the next evolution of it. I think it's like rocket money on steroids for the enterprise. Yeah. Oh, 100%. And I'm sure we'll get there one day. It's actually, I love that you brought that up because we actually had, you know, one of our customers actually bring that up. We're like, hey, can we put that on the roadmap? We're like, absolutely, because that's the direction that we're headed with AI. And again, being able to incorporate AI and start to move forward with all the different opportunities, especially when we talk about agentic AI and having multiple within the enterprise and the organizations and them working together.

14:13There's just so much that can be done in the future. And so, yeah, getting to that point where I can actually go and then start to do that vendor contract renegotiation, again, helps save time, helps save money for companies. It's really important. So if we can help them be as efficient as possible, that's really great. And that means that we're achieving our goal here. And of course, in the future, the vendor will have their AI agent battling with the Altus AI agent to get the best price. And they're probably going to be like, all right, let's just cut to the chase. We're going to, here it is.

14:42But yeah, so the other piece too, there's like the obvious, right? Which is the SaaS license fees, how much I'm paying. I think the part that's hidden in a lot, especially large enterprises, is there's like a lot of humans that if you break it down to their core role and function in the org, they're almost serving the SaaS, all the SaaS products that are out there, or at least a large part of their jobs. I think that's the other interesting part. And this is the big debate is like, AI going to take jobs away or does it free up people's time to do more important things? I'm curious, like the human capital element of this, as you start to identify waste or opportunities within the business, like how does that topic come up?

15:27Because it's a sensitive one, right? Because you're talking about people's jobs, their livelihood, all of those things. Yeah, it is a very sensitive topic. And, you know, it's something that's really important to be had because as we're designing, as we're building, as we're innovating and taking a first principles mentality and approach with everything, we have the opportunity because we are the builders to ensure that as we are designing for the future, we design for the human in mind first. right and that is really important so as long as we're able to stay true to that we're going to be able to build a future where ai is there to help assist right to make things more efficient to make things move with ease so that us the human will have more time to focus on the strategy the creativity the thing that ai won't be able to lean in undo but be able to help in other capacities And I think that's really important.

16:20So I think it's actually all on us to stay responsible on that and ensure that as we're building and innovating for the future, that we have the human in mind first. Right. And how are we doing that? And I just think that when we talked about a little bit earlier in digital transformation, we're also going through a collective global transformation with technology in general. and just it's just a change right and so if we look at history and even like industrial evolution and we go back and see how those moments really change and transform the way that people work yeah there were big changes and so we're experiencing changes as well but i don't think it's necessarily oh okay we're just going to lose jobs but maybe it's transforming different types of jobs so i think that's like the change we're going to see is there's going to be a whole slew of new types of jobs out there.

17:09Ones that we don't even know the titles of yet. Some people may see that, but, you know, and so that's kind of the interesting part and actually the exciting part that actually I see a whole wide range of new opportunities. And so the quicker we can embrace it. And of course, keep in mind as we're designing and innovating, keeping always the human in mind as well, we're going to be in a great spot. It's just also a change and change can be scary, right? Change is different. It's unknown. But I know that if we keep those two things in mind and know about the white space that we're going to have, that we enter this new phase of innovation, there's going to be a lot of new jobs, just different titles, types of jobs.

17:47So that's why I always encourage everyone to go out and test AI, get comfortable with it, because the more that we're all comfortable with it, then it doesn't seem as unknown or maybe as scary. Because now we're a part of the conversation. We're part of the story together. I think, you know, me and you and others are in it every day. So it's in our face. But the large population out there is not. Like, they're not using AI every day. So I think there's that, like, adoption piece. But I totally agree with you. I'll give you an example. I saw this the other day. It was a chart that showed, I believe it was computer engineers, right, leading up to when you kind of had this big dot-com boom and everything that drove.

18:28And those jobs just kind of fell off. But what took off was software engineers because software became the big thing. So it wasn't like, you know, that particular job may have gone away, but those people adapted to new skills and software engineers just blew up from there. And I think it's going to be a similar thing. Like Hattricks AI, the way we're thinking about it is agentic engineers. It changes the role of the engineer where you got to be very deep in the syntax and writing code. and you're kind of leveled up to this orchestrator role, leveraging AI and directing AI. Still a critically important role, still somebody that you want to have experience and understanding of architecture and different patterns and frameworks and things like that, but you're leveraging AI in a sense.

19:13So I think we're going to see that across the board. To your point, it's going to be like new roles and things emerging as time goes on. Yeah, which will be really exciting. I mean, I'm actually very excited about the future and all the things that we are innovating and working through. I think there's just going to be a lot of opportunity. And I think too, like my hot take is, because a lot of people are saying, oh, AI is making things so much easier. We're going to be able to get rid of all this SaaS and do those things. I think it's going to create a whole new group of products and a whole new type of bloat in a sense.

19:46I think that's where companies got to be careful is like you can get into the first principles, the same type of thing where you're chasing the shiny object to do something versus getting to, okay, what's our process? What's our workflow? How is that going to be integrated into it and used effectively? So it's not like some of the examples you're identifying with Altus where we have this product that only two people are using. We have a thousand licenses, you know? Exactly. Yeah. Just be more efficient. Yeah. Go for it. Yeah. With one of the last topics, this is a big one. the SaaS pricing model.

20:23Like how do, what's your thoughts there? Because the traditional approach, like per-seat pricing, the standard in the industry, do you think we're going to see new approaches emerge? Or do you think like per-seat pricing is still a model that we'll see as like AI solutions, you know, become more prevalent? What's your take there on how that all starts to evolve and, you know, the business models in general? Yeah, I think that a lot of changes are going to happen there just because of AI. And so we saw when AI started to come out, a lot of companies were coming out with their agent, you know, agentic AI solutions for businesses.

20:58There was a lot of like add-ons. OK, it's an add-on to charge for this. Right. And then there was actually a lot of backlash. And we've seen that recently where a lot of companies like this is getting very expensive. All of this is so expensive. Right. And companies are really overspending on their tech stacks and they want to really get like a handle on that spend. Like we were talking about that tech bloat and the costs associated with it. I think actually now with AI, actually, I think we're going to enter this moment where folks are really talking about how do we price things a little bit differently?

21:28And I'm starting to see that and hear that in different conversations where different executives are talking about changing up their pricing strategies. Because as we enter this phase of AI, it's like how do we incorporate that more within our solution? And just how we talked about earlier where companies have, you know, their, let's say their SaaS solutions and now they're adding AI versus starting with AI, right? It's a little bit different. But as companies adopt and adopt that more within their solutions and offerings for their customers, I just think that you're going to see a lot of the pricing change just because companies want to be more efficient.

22:03And so I think everyone's kind of thinking through what does the pricing look like in the future. And you can hear a lot of people talk about more of a consumption based pricing model, which is really interesting. And the change there, because there's so much more data flow when we talk about AI, right? AI is all about data. And so I've been hearing a lot more about consumption pricing when it comes to, you know, different pricing models and strategies going to market with AI. So I think we're going to see a lot of changes and I don't know exactly what kind of the next new thing will be in terms of pricey strategy and in those like business models.

22:36But I do think we're going to see lots of transformations there. Like a ton. I mean, even like Sam Altman, I saw him the other day kind of just testing the idea. He bettered out on what platform it was. They're thinking about potentially taking their$20 a month pricing model and moving it to a consumption based option. Right. The idea would be you get like this bucket you can use, but once you use it, you can add to it from there. I agree. I think that's going to be a big one. There's always this, you know, pie in the sky dream of can you get to like outcome based pricing? And I don't know, maybe AI enables that in some way to where you're just paying for the outcome actually being achieved.

23:15I think the other interesting one, back to OpenAI, I don't know if you saw this recently, but they talked about having a$20 ,000 per month option for a PhD level person. It was like 10 ,000 for their software engineer and I think it was like 2 ,000 for their knowledge worker. So it's almost, it reminds me of almost like paying a salary to a human in a sense, but it's for this agentic thing. So I don't know, maybe that's a new paradigm that starts to come up. It'll be, and then the other thing too, think about it this way. it may not just be humans using software anymore like back to agents having access to tools being able to take actions what if i have an agent that needs access to a tool am i paying for a seat for the agent or how does that work i don't know if you've thought down that rabbit hole or if you have any thoughts on that side of it but i love that you brought that up because it's true and then it's like it brings up the question then how do we price things and that's why i think pricing strategies are going to change because then it's going to be maybe a throughput model more where based on consumption, based on the data that's going through on the APIs, then that's going to be more of a pricing, you know, and how it's priced versus a seat.

24:29Because traditionally, we always did a seat because it was always a person. But now if we have multiple agents in an organization and they are running the systems and processes and taking care of, you know, whatever their focus areas are, how do we charge them for that? So that's really interesting. I I love that you brought it up because that's why I really think that pricing strategy is really going to be completely transformed. And I think the consumption model might be more adopted and more popular just because they can actually see what data is being used, what's being consumed, and then price for that.

25:01Because then it doesn't matter per user or per seat anymore. But now it's based on which actually is more aligned with AI and the data that's like going through. Back to that. Yeah. Time to ask. get chat gbt's take on pricing strategy that actually the last yeah like i'm sure we'd come up with some really interesting things too and that's like a bit again back to how you interact with it and how you prompt it and having it come up with novel ideas it can actually do some pretty cool stuff but the last thing i got for you you're obviously a new company an ai kind of native company leveraging ai in your solution like how are you thinking about go to market as like an ai company because I'm sure there's listeners out here out there they're like they want to start a business they see this new like we talked about in the beginning this new blue ocean of opportunity to create new novel solutions and you know new mousetraps to solve existing problems but how do you think about go-to-market when everybody's talking about AI right now I mean heck run the talking AI podcast right AI is everywhere we turn it is like the topic it was actually interesting so earlier matt this year we got to go to the world economic forum and also everything was ai right or at least the track that i was on right so it was around all ai and everything was a agent of ai and i think probably by the end of the week most people were tired of saying that phrase but again it's also just the beginning i say it's always going to be changing every year it's going to transform by next year maybe called something else right uh which is always really interesting especially when we have things like quantum right coming up on the horizon in very quickly.

26:38So we're innovating very quickly. And yet I think these next five years, we're going to see so much transformation in businesses from like a digital transformation, which for me, that's, you know, bringing in AI and what does that look like within the existing larger enterprises globally, right? We're going to see pricing model changes drastically. Like we already see that with OpenAI and they're coming out with so many different new pricing models every day, right? And that's great. They're testing it out. They're seeing what works and every company needs to do the same. And then when it comes to AI and go to market, I just think that if you have a wonderful idea and you're excited to build something and AI helps build that solution faster or helps you go to market faster or is the main, you know, the platform for your solution, go for it.

27:23I think this is, again, a wonderful time to test. And that's how I always say, just test it out. Because what's nice about testing, it doesn't have the pressure of, oh, I need to start this whole business and have a business plan and have this whole strategy. Like, yeah, just go test it out. Give yourself like, I don't know, a month or two and just deep dive into that world. Test out the problem that you're trying to solve for and what it takes to do that. And again, there's great tools like even, you know, ChatGPT. You can go ask and figure out how to code or, you know, build a lightweight no code solution.