PromptCast: The Voice of AI and Security

In this episode, Itamar and Guy Fighel go deeper into the realities of building and maintaining AI systems at scale. Guy explains why data quality still outperforms clever algorithms and why reliability, trust, and observability are becoming the real bottlenecks in modern AI engineering.
The conversation challenges the idea that AI failures are mysterious. Guy breaks down how classic observability practices still matter and where a new layer is required, including dynamic monitoring, real-time feedback loops, and the idea of micro guardrails that guide agent behavior without blocking innovation.
He also discusses how AI is reshaping team structure and velocity. The future, in his view, belongs to smaller and stronger engineering groups with far better tooling and far higher expectations for iteration speed.
Learn more about Prompt Security: https://prompt.security/

In this episode, Itamar sits down with Guy Fighel, a longtime operator, technical founder, and now partner at Hetz Ventures, to unpack the real forces that shaped today’s AI landscape. Guy walks through two decades of data engineering, distributed systems, and the inflection points that changed everything. Hadoop, cloud economics, spot instances, and eventually transformers all shifted what builders could achieve and how fast they could move.
He also brings a rare dual perspective. As both an investor and former CTO, Guy explains why execution still separates companies that ship from companies that stall. He breaks down why understanding budgets, engineering velocity, and technical reality matters more than ever in a world filled with talk about intelligent software.
This is a grounded look at how AI evolved from slow and brittle experimentation into an environment where iteration, scale, and system design decide what is possible.
Learn more about Prompt Security: https://prompt.security/

Cybersecurity isn’t just a corporate issue anymore; it’s personal. From smart homes to connected cars to the devices we carry everywhere, digital risks are woven into daily life. And with AI reshaping the security landscape, the line between digital and physical threats has never been thinner.
While the headlines focus on big breaches, the real dangers often hide in plain sight: overlooked vulnerabilities in consumer tech, quiet privacy erosion, and systems designed for convenience over safety. At the same time, AI is both the hero and the wild card, powering faster defenses while creating new ethical and security dilemmas.
In this episode, Itamar talks with Matt Smyth, cybersecurity leader at CPI Security, about the evolving meaning of “secure” in an AI-driven world. Together they explore the hidden risks inside everyday technology, the future of AI-enabled defense, and how to stay protected in a world where everything (and everyone) is connected.
Learn more about Prompt Security: https://prompt.security/

AI is no longer a side project; it’s at the heart of how modern enterprises operate, innovate, and compete. But as AI moves from experiments to production, the real challenges start to surface: complex infrastructure, data strategy gaps, and a growing need for trust that goes deeper than marketing buzzwords.
Supporting AI at scale isn’t just about model selection. It means rebuilding systems, rethinking governance, and securing every layer of the enterprise. Yet many organizations are still trying to bolt AI onto legacy architectures and hoping for the best.
In this episode, Itamar sits down with Tony Olzak, CTO of Trace3, to unpack what it truly takes to make enterprise AI work in the real world. They discuss the architectural shifts driving AI adoption, the common mistakes companies keep repeating, and how to build trust that actually holds up when AI becomes business-critical.

In this special season finale of PromptCast, James Berthoty, founder of Latio Tech, joins host Itamar for a comprehensive review of AI security in 2024.

The episode examines the year's developments in AI security, from the widespread enterprise adoption of generative AI to the emergence of new security frameworks and governance models.

The conversation explores both technical and strategic perspectives, providing listeners with a thorough understanding of how the AI security landscape has transformed and what lies ahead for 2025.

--------------------------------------------------------

Learn more about Prompt Security: https://prompt.security