Zsolt Balogh: Agentic AI and Enterprise Security | Ep 1252

Summary

The Digital Executive — Episode 1252:
Zsolt Balogh on Agentic AI and Enterprise Security
May 19, 2026 | Hosted by Coruzant Technologies

---

Episode Overview

This episode features Zsolt Balogh, Vice President of Technology Operations & Support at Liferay, discussing the real-world applications and challenges of agentic AI, enterprise security, and the new ISO 42001 AI certification. Zsolt details his career journey, sheds light on how agentic AI is powering complex customer support at Liferay, and shares insights on AI governance, emerging cybersecurity threats, and the future convergence of robotics, nanobiology, and AI. The conversation balances optimism for transformative technology with caution around security and compliance.

---

Key Discussion Points & Insights

---

1. Zsolt Balogh’s Career Journey

Timestamp: 02:20 — 04:05

• Leadership Influences:

  • Zsolt credits early exposure to Stephen Covey’s "Seven Habits" and its emphasis on “beginning with the end in mind” for shaping his leadership and growth mindset.
  • Personal dream: Transitioned from a small Hungarian company to a leadership role at Liferay, driven by a desire to interact with company founders and understand their culture.

• Quote:

  • “For me the turning point was an early exposure to Stephen Covey's seven habits and his son's eight habit. Basically this gave me a foundation on leadership, ownership and continued growth.” (Zsolt Balogh, 02:59)

---

2. Agentic AI in Customer Support

Timestamp: 05:01 – 07:47

• What is Agentic AI:

  • Differentiates agentic AI from generic chatbots: it doesn’t just answer questions, it receives goals and autonomously finds solutions, using tools and breaking complex tasks into workflows.
  • Liferay’s implementation: For customer tickets, an AI agent analyzes context, decomposes tasks, guides human engineers, and provides ticket responses—always with humans in the loop for enterprise-grade caution.

• Impact on Operations:

  • Agentic AI at Liferay achieved a 51% resolution rate for complex, developer-oriented tickets (exceeding expectations for nuanced queries).
  • Paradoxical result: Faster responses led to increased ticket volume, enriching the support process but demanding more human oversight rather than less.

• Quote:

  • “So the agentic AI doesn't just answer your question, but it also acts. You give it a goal and then it will find out the steps, use the necessary tools, and work through the task until it's done.” (Zsolt Balogh, 05:07)
  • “Our responses got so much faster that our customers started to see the value of our support more and they started to open more tickets. So actually it did not replace the support team as a lot of people were expecting, but actually gave us more work.” (Zsolt Balogh, 07:27)

---

3. ISO 42001 Certification for AI Governance

Timestamp: 08:40 – 10:34

• Certification Meaning:

  • ISO 42001 is the first internationally recognized AI management standard; akin to ISO 27001 for security, but targeting AI governance, risk, and compliance.
  • Certification assures customers that Liferay’s use of AI is not just well-intentioned, but externally audited and managed within a disciplined, documented framework.
  • It maps onto emerging regulations like the EU AI Act, thus covering multiple compliance regimes.

• Challenges in Achieving Certification:

  • Not a technical difficulty, but an organizational and communication challenge—aligning all teams to follow standard controls and governance practices.

• Liferay’s AI Hub:

  • Developed to help customers implement similar AI governance, streamlining compliant AI adoption for enterprise clients.

• Quote:

  • “Where an enterprise using Liferay means that our AI is governed and audited in an internationally recognized framework, not just relying on our good intentions…” (Zsolt Balogh, 08:44)
  • “The hardest part… It was not a technical issue, but more like a communication [issue] and how we make sure that our compliance is fast enough.” (Zsolt Balogh, 09:53)

---

4. Looking Ahead — Future Developments and Concerns

Timestamp: 11:07 – 12:17

• Top Concern: Accelerating Cybersecurity Threats

  • Attack surfaces are growing, especially with agentic AI being exploited by malicious actors.
  • The pace of new threats outstrips the speed of conventional defensive innovations.
  • Planning horizons are rapidly shrinking, from years to mere days.

• Most Exciting Development:

  • The convergence of robotics and AI reasoning to serve humanity in unprecedented ways.
  • Anticipates breakthroughs at the intersection of nanobiology and AI within the next 3–5 years.

• Quote:

  • “The cybersecurity threats are accelerating faster than regular defenses can adapt... attackers are scaling up way faster and we really need to be on our toes.” (Zsolt Balogh, 11:07)
  • “I can see a lot of excitement on my end as well about robotics and how merging robotics with AI technologies and reasoning will give genuinely new ways to serve humanity. Looking at three to five years horizon, I think nanobiology converging with AI technical research is where I think the real breakthroughs will be coming.” (Zsolt Balogh, 11:48)

---

Memorable Moments & Notable Quotes

• On Using Agentic AI for Complex Support Needs:
  “These are not just repetitive use cases. And this 51% was really surprising to us… our responses got so much faster that our customers started to see the value of our support more and they started to open more tickets.” (05:45, 07:27)

• On AI Governance:
  “ISO 42k is the first international management standard for AI and basically it's just ISO 27k for AI. It's the same discipline and applied to AI risk. What is great about is that it maps clearly onto the EU AI act…” (Zsolt Balogh, 09:18)

• On The Pace of Security Threats:
  “Planning horizon has shrunk from three to five years to three to five days.” (Zsolt Balogh, 11:27)

---

Critical Timestamps

• [02:20] — Zsolt discusses career influences and pivotal decisions
• [05:01] — Agentic AI explained: practical use in customer support
• [07:47] — Impact on support processes and metrics
• [08:40] — ISO 42001 AI certification: meaning and challenges
• [09:53] — Developing Liferay's AI Hub for customer governance
• [11:07] — Future outlook: cybersecurity threats and AI frontiers

---

Summary Author’s Note:
This fast-paced conversation spotlights how thoughtful AI implementation, organizational discipline, and a pragmatic “human-in-the-loop” attitude can drive innovation and maintain trust in the enterprise sector—while remaining eyes-wide-open about the new risks on the horizon.

Transcript

[Windows 11 College Deal Announcer]

Study and play come together on a Windows 11 PC and for a limited time, college students get the best of both worlds. Get the unreal college deal Everything you need to study and play with select Windows 11 PCs. Eligible students get a year of Microsoft 365 Premium and a year of Xbox Game Pass ultimate with a custom color Xbox wireless controller. Learn [e-mail] studentoffer while supplies last ends June 30th terms at aka mscollegepc

[Indeed Sponsored Jobs Announcer]

when you need to build up your team to handle the growing chaos at work, use Indeed Sponsored Jobs. It gives your job posts the boost it needs to be seen and helps reach people with the right skills, certifications and more. Spend less time searching and more time actually interviewing candidates who check all your boxes. Listeners of this show will get a $75 sponsored job credit@ Indeed.com podcast that's Indeed.com podcast. Terms and conditions apply. Need a hiring hero? This is a job for Indeed Sponsored Jobs Foreign.

[Podcast Host]

Technologies Home of the Digital Executive Podcast do you work in emerging tech? Working on something innovative? Maybe an entrepreneur? Apply to be a guest at www.corazon.com brand welcome to the Digital Executive. Today's guest is Zeult Bollog. Zhe Bolaug is the Vice President of Technology Operations and Support at Liferay, where he leads the company's AI driven innovations across customer support, security and SaaS operations. With a career spanning enterprise technology and digital infrastructure, Zolt brings a deep, practical approach to some of the most complex challenges facing modern software organizations, from scaling support operations to defending cloud platforms against increasingly sophisticated threats. A sought after international speaker, Zolt regularly shares insights on the real world applications of AI, cutting through the hype to focus on what actually works in production environments. His talks cover topics ranging from agentic support systems to operationalizing machine learning at enterprise scale, making him a trusted voice at the intersection of AI, security and customer experience. Well, good afternoon Shult. Welcome to the show.

[Zolt Bollog]

Thank you.

[Podcast Host]

You're very welcome my friend. I appreciate it and making the time you're hailing out of Budapest, Hungary today. I'm in Kansas City, so appreciate you traversing all these time zones and calendars to get here. So thank you again and jolt. Let me jump right into your first question. You started your career as a system admin, joined liferay as the European Support Team lead, and have since risen to Vice President of Technology Operations, earning CISO of the Year recognition along the way. What were the pivotal moments or decisions that shaped that journey? And did you ever imagine you'd be leading AI and security Strategy at this

[Zolt Bollog]

scale started my career in hands on technical roles as you explained. And for me the turning point was an early exposure to Stephen Covey's seven Habits and his son's eight habit. Basically this gave me a foundation on leadership, ownership and continued growth. And then those values actually match really well to my physical religion and culture. The beginning with the end in mind. Habit actually worked for me before I even knew about it. I had a dream that when I was working for a small Hungarian company that I want to work for life free. At one point in my life I was working for the community and I really wanted to meet the founders and talk to them and understand how they have built this beautiful company. And I still cannot believe that I didn't just get a chance to work for this company, but also can lead these areas.

[Podcast Host]

Thank you, really appreciate that. You know I talked to a lot of technologists here on the podcast, of course, a lot of CISOs and VPs of technology etc, but I like the fact that you jumped in and really did some reading and I like we've all, I think a lot of us have read this book. Stephen Covey's book on the seven Habits have highly effective people and I'm glad that was influential in your career as well. And you're very curious. You got out there and started to meet, interact with people along the way, building those relationships as your career blossoms. So thank you and Jolt. You've been championing agentic AI systems to transform how support teams operate at liferay for executives who are still wrapping their heads around what agentic actually means in practice. Can you walk us through a real world example of how these systems are changing the customer support experience?

[Zolt Bollog]

We have started to work on improving the customer experience with agent AI. Basically before it was an end agency. So the agent decay doesn't just answer your question, but it also acts. You give it a goal and then it will find out the stats, use the necessary tools and work through the task until it's done. So what we have built at Lightframe for our customer support team was that when a customer ticket is open, then an agent reads it and based on the complex information available data, it will start breaking down the tasks into multiple different workflows and give a response on the ticket to help the support engineer. So this is a very cautious way of course using agentic fireflows because we want to make sure that our enterprise customers are served well. So we still have human in the loop responding, although we are exposing more and more of these to our customers. So our Agentix system is of course using and searching the documentations and known issues and referencing for example, our security vulnerability documents. And at the same time, it can also do very complex things. For example, if there's a performance issue, we have made a very special agent which can understand the most complex performance issues and give very specific steps on how to resolve those. So what did it mean for the customers that we are doing faster triage? We measured actually using AI the back for one service agent, and it can give actual proper ticket resolution responses in 51% of the cases. So at Liferay, we are supporting customers who have very complex use cases because our customers who are opening the tickets are also developers because they're building on our own on top of our platforms. So these are not just repetitive use cases. And this 51% was really surprising to us and we were to narrow it down. But it turns out that most of the tickets don't have enough context at the beginning. So we are working on to improving the quality of the issue reports. At the same time, the surprising result was that our responses got so much faster that our customers started to see the value of our support more and they started to open more tickets. So actually it did not replace the support team as a lot of people were expecting, but actually gave us more work. But it also brought more value to our customers, which they do really appreciate.

[Podcast Host]

That's amazing. Thank you so much. And I appreciate that you talked a little bit about agentic AI. You impact some of the capabilities obviously of what an agent can do. It's obviously more than just answering questions. It can act and do tasks, as you mentioned, including breaking down multiple tasks into workflows. You noted that it's important to keep the humans in the loop, which we talk about a lot here on the podcast. But I like that you have created a special agent to handle very specific and complex tasks, especially for your developers, because generally that's who your customer base is. So thank you. And Schulte. Liferay recently became one of the first companies globally to achieve ISO 42001 certification for AI. What does that certification actually mean for an enterprise considering liferate? And what was the hardest part of getting there?

[Zolt Bollog]

Where an enterprise using life right means that our AI is governed and audited in an internationally recognized framework, not just relying on our good intentions on how we are deploying AI, we are getting documented controls on how we are being deployed and monitored and improved actually. And an external body has verified that we actually did the work, which we are telling everyone. So ISO 42k is the first international management standard for AI and basically it's just ISO 27k for AI. It's the same discipline. And applied to AI risk, what is great about is that it maps clearly onto the EU AI act, for example, and any emerging sector rules. So it actually satisfies multiple regimes at once. The hardest part was honestly is not the technical controls. It was that there's a pretty big fraud from everybody's side to apply new AI technologies. And making sure that we can follow the same controls, follow the same playbook and making sure that all departments are doing this was the biggest concern. So it was not just. It was not a technical issue, but more like a communication and how we make sure that that our compliance is fast enough. And to support that, we recognized that several organizations will have some limitations on how they can give access to AI data. So we actually figured out that we can build a product which we call AI Hub, which helps customers to bring the same governance and structure to their AI use. So they will have a faster go to market basically with their AI tools because then they can have more controls around the systems.

[Podcast Host]

Thank you, really appreciate that. And I like that you have built in that strict AI governance with your platform, but you can extend that to your customers and make sure they are also provide those guardrails and appliance as you talked about. But I think that's really cool that certification, that ISO 42k basically means you're doing what you say you can do and that validates that. So I appreciate that. And Zolt, last question of the day. Looking out maybe three to five years, what's the development you're most excited about and what's the one that keeps you up at night?

[Zolt Bollog]

I'm working a lot with security, so let me start with the concern first. So the cyber security threats are accelerating faster and then regular defenses can adopt. We are seeing supply chain issues every single day, even this morning. And there are new attack surfaces coming, for example from agentic AI attackers are scaling up way faster and we really need to be on our toes and actually continuously improving how we are defending. And this is not three to five years. Basically half jokingly, our planning horizon has shrunk from three to five years to three to five days. On the excitement though, I can see a lot of excitement on my end as well about robotics and how merging robotics with AI technologies and reasoning will give genuinely new ways to serve humanity. Looking at three to five years Horizon, I think nanobiology converging with AI technical research is where I think the real breakthroughs will be coming.

[Podcast Host]

Thank you. Really appreciate that. And I, like you talked about agentic AI and the advanced reasoning that is starting to do now. And of course, I believe if you look at the glass half full, we can see where AI can certainly improve the way we work, the way we live, improve humanity all around. But again, we got to make sure those guardrails are in place and there's people still in the world that are leveraging AgentIC AI for for the wrong reasons. And that's why cybersecurity keeps you and me both up at night, because these threats are evolving much faster, especially with the advancement of these agents. So I really, really appreciate that. And Zolt, it was just such a pleasure having you on today, and I look forward to speaking with you real soon.

[Zolt Bollog]

Thank you so much. It was great talking to you.

[Podcast Host]

Bye for now.