AI Security, ROI and Autonomous Agents, How to Scale Safely with Chris DeNoia

Speaker A: Built to Scale with Reem Benshaw. This is Built to Scale, the podcast for B2B founders who are playing the long game. I'm your host, Reem, and every week we are exploring the mechanics of sustainable growth with the founders and executives who are defining the industry. No fluff, no growth hacking gimmicks, just the blueprints for building a solid, scalable tech company. Welcome to the show. This episode is brought to you by PartnerPropel. Hiring humans to do robot work is a waste of talent. Partnerpropel transitions you to an AI first marketing and sales company. We built the infrastructure that automates the grunt work. So your team focuses on one thing. Closing deals, stop renting results, and build an engine that you own. Visit partnerpropel.com to get started. Today we have a special guest joining us, Chris De Noia, fractional CISO and cio. Chris, thank you so much for joining us today.

Speaker B: It's a pleasure to be with you, rem. Thank you.

Speaker A: So tell me more about yourself.

Speaker B: I appreciate it. So as you articulated, I'm a fractional executive in the space, uh, serving as CIO and CSO for organizations, helping them transform in this world of AI, but also doing it securely at the same time. And I'm having a great time doing it. We're hitting some phenomenal achievements and I think the pipeline is looking very well so far.

Speaker A: Very good. I think we talk a lot about a lot of, uh, about AI and all those sorts of kind of trends happening in the tech space, but we rarely talk about the security side of it, which is probably the most important part, let's be honest. So I'm kind of excited to kind of talk to you and have you on the podcast and dig into some, some topics that are going to be very relevant for a lot of people that we don't discuss enough. So happy, um, to have you here. So let's talk about hiring in a market like the one that we have today that is very selective. What do you think is one specific skill that separates an AI enthusiast from a hire that a CEO will actually feel safe betting on today?

Speaker B: You know, CEOs don't need prompt engineers. They need risk aware system architects. So here's what I mean. The market is flooded with tourists right now. Anyone can prompt, but a CEO really needs someone that is an agent literate, someone that can identify workflows, someone that can identify the risks, assess the risks, measure the impact. Enthusiasts, they, they, they play with chatbots. The system thinkers are the ones that actually scale an organization. And I think Ultimately, for an executive, those are the individuals you want on board for the next era that is now upon us. It's not even yesterday. Uh, it is now today.

Speaker A: Absolutely. Um, and in terms of kind of the execution, right, um, now that you feel like the hype died down, what do you think is the most common reason why you're seeing promising AI projects actually fail to generate a, ah, return on investment?

Speaker B: Such an important question from my perspective of empirical experience, what I'm seeing thus far, as well as industry trends and reports. This is business leaders are trying to boil the ocean instead of boiling a single cup of water. Uh, they want a massive transformation, but they over scope and a lot of times that delivers disappointing results. The winners in 2026, those are going to be the exact opposite. Right? They're going to find that one painful, repetitive workflow, they're going to automate it. They don't start large, they start small, they start measured, and then they compound those wins. Now in the age of AI, we move very quickly nowadays, those compounding wins can really take off rapidly. Very, very. So start small, start measured, compound your wins and before you know it, you'll have your transform.

Speaker A: I love this. I do think this is a, probably a more realistic and sound approach than as you mentioned, trying to boil the entire ocean. Um, how, how do you think this is going to drive return on investments? How do you essentially look at ROI when it comes to this kind of approach? Uh, kind of starting small and building from there.

Speaker B: So ROI is derived through measurement, of course, and calculation. We start to think about KPIs. A good project is always going to have those attached KPIs that you're measuring. So that way you can understand whether you're really hitting those goals that moving the needle for the organization. So whatever the executive wants to choose, wherever they start, clear metrics are very important to demonstrate the fact that you are capable of delivering those compound wins. And that itself allows you to present to the CFO, or maybe you are the CFO, or maybe the CEO. Either way it allows you to translate those KPIs into that ROI. But if you do not measure it, point back to the measure for the organization's projects, you're not going to have that ROI measurement in general.

Speaker A: That makes sense. And I feel like we're, we haven't talked enough about ROI overall in this day and age where AI is becoming essentially, uh, essential to, to companies, to most people's roles and is really kind of requiring large budgets right now. So I love the fact that you're kind of reiterating what, what matters the most, which is what are the KPIs, which sounds essentially like a fundamental thing and principle. So thank you for sharing your thoughts around that. So on security versus speed, right now everyone is obsessed with executing fast right to drive as much money and revenue as possible. Where is that speed creating the biggest security blind spot for companies right now?

Speaker B: So we're evolving into this agentic workforce era, right? Yesterday's risk was a, uh, chatbot saying something inaccurate. Today's risk is different. Today's risk is autonomous agents wiring a million dollars to a hacker. Speed is creating a massive blind spot right now, and that's titled shadow AI. So when AI moves from suggestion to education or technically, uh, execution. Excuse me, we need to move from static policies to dynamic operations. So think layered safeguards. So development, runtime, testing, and then observation and government governance though these are all the layers in which are necessary as we're moving into this agentic.

Speaker A: Do you think companies are ready for this? I mean, uh, you kind of mentioned some alarming potential scenarios. Right. And I think that again indicates that we're not there in terms of being able to create enough privacy guardrails and insecurity guard like grails. Especially this day and age where we're, we're building AI agents and such. Where do you think companies are at in terms of, of, of protecting themselves?

Speaker B: Yeah, um, I think most organizations are currently on the journey towards a maturity level and you either have pulled the trigger and you have started or you have not. Uh, it depends on which report you're reading and what they're seeing in optics. That being said, along that journey, somewhere, somewhere along the line, you eventually discover the fact that if you're going to bring in these autonomous agents, which I, I perceive them as more of a, uh, an independent body. It's another employee. And the way that I, I perceive it that way allows me to then articulate in the fashion of more human resource type of operations than is about anything else. And the reason for that is because these entities move faster than humans do and they don't sleep either. And if we're going to give them agency, if we're going to give them tools, that means they have the ability to make positive impact and catastrophic results at the same time, either way. So we need to think of it from a human resources perspective. So that way we have something tangible to go back to. So to answer your question, organizations need to get there if they're going to eventually evolve into this agentic Workforce, which has a great amount of potential for organizations depending upon what you want to do, when you want to do it

Speaker A: and how you want that makes sense. I love the way you're, you're looking at AI agents like just autonomous bodies. And I do love your perspective because it's definitely real and it helps us understand a little bit more what's at stake stake there in terms of not just a positive impact, but the catastrophic potential impact. So thank you for sharing your insights here. So in terms of remote work, with remote work and roles essentially disappearing, do you think AI implementation is just too collaborative and high stakes to be done effectively by a distributed team?

Speaker B: I do not. Uh, physical geography is irrelevant. Forcing engineers to sit at a cubicle, shared a cubicle won't make AI any smarter. Uh, we're entering again into that agentic workforce era. The highest performing teams of 2026 are, aren't sitting next to each other at the same desk. They're building the best digital colleagues that the organization can possibly dream of right now in this current state. So you don't scale by mandating a badge swipe. You scale by engaging in workforce with superhuman capabilities just for you. The best colleagues with the best capabilities designed specifically for what you want to do. That is not even a 10X. That is a X multiplier situation. And it has nothing to do with swiping a badge and sitting at a cue ball. It will drive the needle for you much more than physical capacity.

Speaker A: It's amazing. Sometimes I have to pinch myself and realize that this is not really a sci fi dream that we're living in, but the reality with the way you're speaking about, you know, building essentially AI colleagues, I mean we're seeing it happening, but the way you're, you're describing it is, is kind of making it even more real. So again, really interesting perspective and futuristic. But I think the future is, is, is closer than, than we think. So let's move on to strategy. Right. For a company that has plenty of AI ideas but essentially zero execution yet, what is the first move they should make to prove that they aren't just chasing the trend?

Speaker B: So my recommendation is to at this point in time, stop strategizing and start deploying low risk, measurable technology, in this case agents. So the age of vibe adoption, depending upon what vernacular you want to track, is technically dead. We need to stop with the 10 year roadmaps, the committee meetings, and so on and so forth. You want to start with that one highly repetitive task that is burning human labor that value asset that really makes you shine as a business. You want to attach those clear metrics that we were talking about earlier, and you want to build the simplest agent that will solve that exact problem. And that's very important. Do not overbloat, do not over scope. The simplest agent for eloquent purposes will do the job, measure it, track it, govern it, and compound those wins. And again, because we're moving so quickly, this technology allows you to compound quicker than ever before. You will have compounding wins quicker than ever before. But start with that one particular thing, wherever it is in your business operations, start there and compound.

Speaker A: And I love that because I, I speak to a lot of founders and entrepreneurs and a lot of them are very overwhelmed with all the tools available, all the use cases for AI, and sometimes that just leaves them in a place where they are kind of an analysis paralysis situation and they're just not sure where to invest, where to start. But the way you're describing this is so much simpler and realistic and reasonable. Just start with one task that is high in human labor and essentially automated, and it compounds from there. I do think it is simple, but it's probably the best advice I've heard in terms of AI adoption for companies, uh, in the tech space. So really appreciate your, your wisdom here and perspective. Cool. Well, I mean, this was a great podcast. Thank you, Chris, for joining. For everybody else who wants to connect with you, learn more about what you do, what would be the best way to do that?

Speaker B: Absolutely. So I'm on LinkedIn, as much of the rest of the world is at the moment, by all means, please reach out. I would love to connect.

Speaker A: Awesome. Um, thank you again, Chris. For everybody else who listened to the podcast, if you enjoyed this episode, make sure you're giving us a follow, give us a, uh, review, or share this episode with someone else who would benefit from hearing it. Thanks again, Chris, and thank you everybody else for listening and see you next time.