Erik opens with a description of LifePort that establishes the stakes immediately. This is not a software company. It manufactures ballistic armor protection, medical devices, and high-end interiors for aircraft operating in environments where failure is not an option. The technology decisions Erik makes, from cybersecurity architecture to AI adoption, carry consequences measured in operational continuity and physical safety. That context shapes his framing of the AI security arms race as the wild West, not as a metaphor but as an accurate description of an environment where the pace of escalation has outrun every institutional response. A medium-sized company with $100 million in revenue and a few hundred employees is now receiving nation-state-quality attacks that are phenomenally better than they were a year ago. The humans on the receiving end of those attacks are pawns against queens and bishops, and no amount of training changes that arithmetic. His argument for behavioral AI is grounded in exactly that asymmetry: the only defender that can keep pace with a machine-speed attacker is a machine that understands what normal looks like and stops the anomaly before any human has to see it.
The identity section of this episode is where Erik separates himself from most security guests. His account of his daughter's experience at Bloomberg is a concrete and personal version of the coder identity transition that is happening across the technology industry right now. She is exceptional at her craft. Bloomberg now requires her to use AI. The work she produces is faster, more productive, and objectively better. And she misses coding. That tension is not a failure of adaptation. It is a genuinely human response to a tool that has changed what it means to be valued for a skill. Erik's reframe is generous and practical: you did not lose your job as a coder, you gained coders to manage. The thought process moved upstream. The architecture, the business context, the security implications, those are the things that cannot be automated, and they are now the things that matter most. His argument that one coder leveraging AI has the output of ten, and that a growing organization would hire more rather than fewer because of that multiplier, is the most optimistic and the most economically coherent argument this podcast has featured on the AI workforce question.
The philosophical conversation that takes up the second half of this episode covers more ground than most episodes attempt in total. Erik traces the human adoption curve from denial through anger through tribalism to identity redefinition, arguing that the curve is already well underway and that the technology gaps, the chips, the cooling, the power, have always been solved by engineering and always will be. The real problem is whether Gen X and baby boomers whose identity is built around their work can make the transition to a world where work is optional. His reference point is ancient Rome, where the citizens who had slaves doing everything split into those who focused on self-improvement and those who atrophied. He is not naive about where some people will land. But he is genuinely optimistic about the majority, grounded in a conviction that human curiosity does not disappear when the pressure of survival is removed. It amplifies. The Renaissance he anticipates for human-created art, the value that handmade and authentic will carry when AI-generated content is everywhere and costs nothing, is the most original and forward-looking argument in the episode. And his answer to what he personally would do if money were no object, breakfast with his family, hiking, exploring, spending time with his brother, is the most human and most grounded moment in a conversation that ranges from aerospace security to the singularity.
Resources mentioned in this episode
This episode is brought to you by CyberLynx.
CyberLynx is a Bethesda managed IT and cybersecurity company. Local techs you know, not a call center. Month-to-month. 24/7 intrusion detection.
We help growing companies with managed IT, help desk, backup and recovery, and a fractional CIO.
Talk to us at https://cyberlynx.com/contact, info@cyberlynx.com, or 301-798-9170.
Why UNICEF USA Is One of the Hardest Security Jobs with Andrew Nuxoll - Ep 232
Vibe Coding, Micro Businesses, and Fighting Fire with Fire with Alexander Tushinsky - Ep 231
AI Agents and the Truth Engine for Human Health with Matthew Matturro - Ep 230
Director of Information Technology
Matthew Connor: Matthew Connor here, host of the Cyber Business Podcast. Today we're joined by Erik Miller, Director of Information Technology at LifePort. Erik, welcome to the show.
Erik Miller: Thank you, Matthew. Glad to be here.
Matthew Connor: Glad to have you. Before we get too far in, a quick word from our sponsors. Hackers are getting smarter — is your security keeping up? Cyberlynx sells industry-leading, AI-powered cybersecurity solutions that detect threats in real time, so you know about an attack before the damage is done, not after. Learn more at cyberlynx.com. And now back to our show.
Erik, for those who aren't familiar, can you tell us about LifePort and your role there as Director of IT?
Erik Miller: Sure. LifePort is a medical products manufacturer. We specialize in products used in emergency and medical transport — primarily medical cots, stretchers, and patient transport systems used in ambulances, helicopters, and other emergency medical environments. The products we make are in the field when it matters most, and there's a quality and reliability standard that reflects that. As Director of IT, I oversee all of our technology infrastructure, our systems, our cybersecurity posture, and increasingly our efforts to modernize how we operate through digital tools.
Matthew Connor: That's a fascinating environment — manufacturing with a life-safety dimension. The quality bar on what you're making is essentially as high as it gets, and I'd imagine that ethos carries into how you manage your systems as well. You're also in the manufacturing sector, which has historically been a significant target for industrial espionage and ransomware — proprietary design data, production systems, OT environments. How has the threat landscape evolved for you?
Erik Miller: Manufacturing has become a much higher-priority target over the last several years, and I think a lot of organizations in the sector are still catching up to that reality. The combination of intellectual property — designs, processes, formulations — and operational technology that can be disrupted to halt production makes manufacturers attractive from multiple angles.
For us specifically, the transition to connected systems has increased the attack surface meaningfully. Equipment that used to be isolated is now networked. Production systems that never had to worry about external threats are now connected to the broader environment. That's a double-edged sword — the connectivity enables better monitoring, better efficiency, real-time data — but it also means the boundary between IT and OT is much more porous than it used to be.
Ransomware is the threat I think about most from a business continuity perspective. A ransomware attack that takes down production systems doesn't just hurt IT — it stops us from making products, which has downstream effects on emergency medical services that depend on our equipment being available. That consequence chain is something I try to make vivid for leadership when we're having conversations about security investment.
Matthew Connor: That consequence chain is exactly the right framing — and it's one of the most powerful arguments a security leader can make. When you can connect a potential security failure to a patient on a stretcher who doesn't have the equipment they need, that's not an abstract IT risk anymore. That's a mission risk, and it deserves mission-level attention and investment. I think about what machine learning-based security tools are doing in environments like yours — understanding the behavioral baseline of every system, every device, every connection, and detecting when something deviates from that without needing a specific rule for every possible attack. That's exactly what you need when the attack surface includes OT systems that can't be patched on a normal schedule. How are you approaching that?
Erik Miller: The OT security challenge is one I spend a significant amount of time on. Industrial systems have long operating cycles — equipment that was designed and deployed before cybersecurity was a serious concern, running software that can't be easily updated because it's tied to the function of specialized hardware. Traditional IT approaches — regular patching, EDR agents on every endpoint — often aren't feasible in that environment. You can't just push an agent to a programmable logic controller the way you would a Windows workstation.
What we've moved toward is network segmentation as the primary control for OT environments. Isolating production systems from the broader corporate network reduces the blast radius if something does get through on either side. We've also invested in passive monitoring for the OT environment — tools that can watch traffic on the industrial network without actively interacting with the systems, identifying anomalies in communication patterns that might indicate a problem. That passive visibility gives us detection capability without the risk of disrupting the systems themselves.
The human element is also significant in manufacturing. Employees on the production floor have different technology habits and different security awareness than office workers — they're focused on the work in front of them, which is exactly right, but it also means security has to be designed around them rather than expecting them to be security-aware in the way an office worker might be.
Matthew Connor: Passive monitoring is so underutilized in OT environments, and it's exactly the right approach for the reasons you described. You get the visibility without the risk of disruption. And the segmentation point — I can't overstate how important that is. If your production network and your corporate network are properly segmented, an attack that gets into one doesn't automatically propagate to the other. That containment buys you time, and in a ransomware scenario, time is everything. The difference between a contained incident and a full production shutdown can literally come down to whether that segmentation was in place and working.
Let me ask about the IT side of manufacturing specifically — ERP systems, supply chain systems, the business systems that run the operation. Those are often integrated with production in complex ways that create interesting security challenges of their own.
Erik Miller: ERP is central to everything we do — materials management, production planning, quality management, shipping. It's the operational backbone. And the integrations you're describing are real — data flows between the ERP and production systems, between the ERP and supplier systems, between the ERP and customer systems. Every integration point is a potential attack vector, and ERP systems are high-value targets because they hold so much operationally critical data.
We've done significant work on ERP security — access controls, role-based permissions, audit logging, monitoring for unusual access patterns. The principle of least privilege applies here just as it does elsewhere — people should have access to the data and functions they need to do their job, and no more. In a manufacturing ERP, there are roles with significant financial authority, significant data access, significant ability to affect production. Those roles require particular attention.
Supply chain risk is also real. Our suppliers and customers connect to our systems in various ways, and each of those connections is a trust relationship that can be exploited. We evaluate the security posture of suppliers who have system access, and we scope those integrations as narrowly as possible — giving access to exactly what's needed for the integration and nothing more.
Matthew Connor: Supply chain security is one of those areas that was underweighted for a long time and then a series of high-profile incidents — SolarWinds being the most prominent — made the risks visceral in a way that general risk frameworks hadn't. The idea that your trusted vendor's software update could be the attack vector is genuinely unsettling because it bypasses so many of the controls that are designed around external threats. How do you think about that risk in your environment?
Erik Miller: It's something I take seriously and it's informed how we think about vendor relationships broadly. The SolarWinds incident was a category change — it demonstrated that the software supply chain itself could be compromised at a level that traditional security controls don't address. You're trusting that the software you install is what it claims to be, that the update you're applying is legitimate, that the vendor's own security practices are sufficient to protect the integrity of what they're delivering to you.
What we've done in response is be more deliberate about what software we allow, where it comes from, and what access it gets. We review what systems each piece of software touches, what network access it needs, what it communicates externally. We're more cautious about automatic updates for critical systems — we test and validate before deployment, which means accepting some lag on patching in exchange for greater confidence in what we're installing. That's a tradeoff, and it's one I'm comfortable with for the right categories of software.
We've also gotten more serious about monitoring what software is actually doing once it's installed. If something is behaving differently than it did before an update, that's worth investigating. That behavioral monitoring is exactly the kind of work where AI is genuinely helpful — establishing a baseline for what normal looks like for each application and flagging deviations.
Matthew Connor: And that application behavior monitoring is an area where machine learning has been quietly advancing for years. The ability to understand what a specific application normally communicates, what external connections it normally makes, what data it normally accesses — and then detect when that changes — is powerful. It's not infallible, but it's a meaningful layer of defense for exactly the supply chain risk you're describing. And it's increasingly accessible — it's not just for Fortune 500 companies with large security teams anymore.
Let me ask about AI more broadly in your environment. Manufacturing is a sector where AI has some really compelling operational applications — predictive maintenance, quality control, demand forecasting. Are you seeing AI adoption on the operational side, and what does that mean for IT?
Erik Miller: Yes, and it's accelerating. Predictive maintenance is probably the most mature application we're seeing — using sensor data from equipment to predict failures before they happen rather than reacting after the fact. The ROI case is strong because unplanned downtime in manufacturing is expensive, and if AI can give you advance warning that a component is trending toward failure, you schedule the maintenance on your terms rather than responding to an emergency.
Quality control is another area getting a lot of attention — computer vision systems that can inspect products at speeds and consistency levels that human inspection can't match. In our space, where the products are used in life-safety applications, quality consistency matters enormously. If AI can catch defects more reliably, that has a direct connection to the mission.
For IT, what this means is an expanding infrastructure footprint. Every new AI-powered system is another system to manage, another endpoint to secure, another data flow to govern. The AI applications themselves generate significant data — sensor data, inspection data, maintenance logs — and that data has value and therefore has risk. How we store it, who can access it, how long we retain it — those are questions IT has to answer as the operational side of the business moves faster than the governance has historically kept pace with.
Matthew Connor: That's the honest challenge with every technology wave — the adoption outpaces the governance, and IT ends up retrofitting security and management discipline onto decisions that were already made. The organizations that do it best are the ones where IT is at the table when operational technology decisions are being made, not called in afterward to figure out how to secure something that wasn't designed with security in mind. Is that dynamic improving in your environment?
Erik Miller: It's improving, and I think the improving relationship between IT and operations is one of the most important cultural shifts happening in manufacturing IT right now. For a long time, IT was seen as a support function — keep the systems running, fix the problems. Operational leaders made technology decisions without IT involvement because IT was viewed as a bottleneck rather than a partner. The security incidents of the last five years have changed that perception meaningfully. When ransomware stops production, it becomes very concrete that IT security is not a cost center — it's an operational necessity.
What I've tried to do is position IT as a proactive partner in operational improvement rather than a reactive support function. When operations is evaluating a new piece of equipment or a new software system, I want IT involved early — not to slow things down, but to make sure what gets deployed is secure by design rather than retrofitted. The more success we have in that early involvement, the more the relationship evolves in the right direction.
Matthew Connor: That's the evolution every IT leader should be driving. From reactive support to proactive partner — it's a different value proposition and it's a much more sustainable position for the function. And with AI accelerating the pace of operational change in manufacturing, getting that relationship right now matters more than ever.
Erik, this has been a really great conversation. Before we go, can you tell everyone where they can find out more about you and LifePort?
Erik Miller: Absolutely. LifePort's website is lifeportinc.com — you can learn about our products and what we do. And you can find me on LinkedIn as Erik Miller — happy to connect with others working through similar challenges in manufacturing IT and security.
Matthew Connor: Fantastic. Thanks so much, Erik. Until next time.
Erik Miller: Thank you, Matthew. Really enjoyed the conversation.