Cyber Sec eSummit 2025

Adam has worked on numerous projects at both international and domestic enterprise companies in leadership, architectural, and expert roles (HP, Nokia, Vodafone, EPAM, United Consult). His main areas of expertise include Observability, Performance Engineering, and Quality. Beyond client projects, he has also been a speaker at international and domestic conferences such as HUSTEF, TestCon, and Mndwrk events.

Enhancing Cybersecurity with Continuous Attack Simulation and Threat Exposure Management

As cyber threats become increasingly sophisticated, organizations need to shift from reactive to proactive cybersecurity strategies. Attack simulation technologies offer a powerful approach to continuously validate security measures, identify vulnerabilities, and enhance overall resilience against potential threats.

The adoption of attack simulation and continuous threat exposure management technologies is a critical step toward building a resilient cybersecurity framework. These tools empower organizations to move beyond traditional defenses and maintain a strong security posture in an ever-changing digital world.

Omri is the Co-founder and leading G2M efforts of Twine, whose mission is to build the next generation AI workforce. Previously Omri was a Partner at Dell technologies Capital, a Tier 1 VC firm. His career also includes significant stints at Grove VC and Claroty. As a seasoned entrepreneur, Twine marks his second foray into founding a company. Omri earned a degree in Electrical Engineering from the University of Tel Aviv. He leverages his technical expertise and practical insights to drive digital transformation. Omri maintains close ties with a network of cyber leaders to foster innovation and share insights.
‍

From Reactive to Proactive: AI Agents in Identity and Access Management

IT teams chase orphaned accounts, duplicate identities, and systems that never sync. They fix one problem and three more appear. This is reactive work. It kills productivity and creates risk. AI agents change this. They hunt problems before problems hunt you. They clean data while you sleep. They prevent the fires instead of fighting them. Real deployments show 80% less manual cleanup. Better compliance. Happier users. This presentation shares five practical lessons from real-world deployments of AI Agents for identity data remediation, covering both technical approaches and organizational considerations for gaining stakeholder buy-in. Attendees will learn actionable strategies that align with Gartner's recommendations for modern identity governance.

Leader and team member of MWT Solutions, ManageEngine solutions  expert with experience from business process supporting software solutions and IT security.

Zero Trust and Automated IT Security in Practice – with ManageEngine Solutions

The Zero Trust model and automated IT security are becoming increasingly important in the modern enterprise environment, especially with the rise of hybrid and cloud-based systems. This presentation demonstrates how the principles of Zero Trust can be implemented using ManageEngine products (such as ADManager Plus, Log360, PAM360, Endpoint Central), and how to automate privilege management, logging, incident handling, and compliance reporting. Through practical examples, the talk will show how to reduce risks caused by human error, accelerate response times, and ensure continuous compliance.

The presentation answers the following questions:

- How can the Zero Trust model be implemented with ManageEngine tools?
- What automation options exist for privilege management, logging, and incident handling?
- How do ManageEngine solutions help meet compliance requirements (e.g., GDPR, ISO 27001)?
- What are the best practices for integrating ManageEngine tools into a modern, hybrid enterprise IT environment?

Dávid has a background in software development, and security has always played a key role in his work. He has been involved in both successful and unsuccessful ventures. The development of OpenShield began at the end of 2024, providing a security layer for LLM models.

Why is it difficult to protect LLM models and AI agents? Prompt Engineering vs. Prompt Attacks: The Battle for Control Over AI Minds

In addition to LLM models, MCP servers have also emerged. Their proliferation introduces new security risks, as LLMs are just as vulnerable to attacks as other systems. The main challenge lies in the diversity of attack types. Traditional security solutions alone are no longer sufficient. Moreover, MCP systems can provide direct access to corporate data, further increasing the risk.

This presentation will address the following questions:

• Why is it important to protect LLM models?
• What is prompt injection?
• What challenges must be faced?

‍

Quantum Threats, Real Solutions – The Future of Secure Data Sharing

The rapid advancement of quantum computers could fundamentally change our current understanding of data security. Most encryption methods widely used today may become vulnerable within a few decades, making the research and development of quantum-proof solutions essential. This presentation will review the theoretical and practical challenges of quantum-proof encryption, introduce the principles of post-quantum cryptography, and discuss methods already available for achieving future-proof data transmission. The goal of this presentation is to highlight the importance of quantum security in data sharing and communication through practical examples.