Operations | Monitoring | ITSM | DevOps | Cloud

Real-time camera deepfakes are no longer science fiction. High-fidelity, AI-generated impersonation may be advancing quickly — but that's not the only AI risk financial services companies should be thinking about. In this episode of AI Can Do What Now?!, Lisa Jones-Huff, director of security solutions architecture at Elastic, sits down with ethical hacker Freakyclown (FC) to explore what is technically possible today with AI, where reality still falls short of the hype, and what security teams should be worried about.

What is the most immediate risk financial services companies face today? AI-enabled social engineering is already accelerating real-world attacks. Scale, personalization, speed, and automation are lowering the barrier for attackers while making fraud detection more complex for defenders. In this episode of AI Can Do What Now?!, Lisa Jones-Huff, director of security solutions architecture at Elastic, is joined by ethical hacker Freakyclown (FC), and principle solutions architect Joe Murin to explore what is actually happening right now — beyond the hype.

Building AI applications often means juggling multiple models, scattered datasets, and version chaos across local systems. But what if you could bring it all together — securely and efficiently — in one place? In this walkthrough, Shibam Dhar, DevRel Engineer at Harness, demonstrates how Harness Artifact Registry makes it easy to manage and govern your AI/ML assets — from models and datasets to prompts and agents — with built-in support like Hugging Face and generic registry types.

Spec-driven development with AI coding agents could change how you build software. In this GitKon 2025 talk, Erik Hanchett, Senior Developer Advocate at AWS, breaks down why AI coding assistants perform dramatically better when they start with structured specifications instead of raw prompts. If you've been vibe coding your way through complex features and wondering why your AI keeps going off the rails, this is the video for you.

Most enterprise SaaS products, like Komodor’s Autonomous AI SRE Platform, require installing a remote agent on the customer’s infrastructure, which varies significantly from one organization to another, in terms of architecture, configurations, permissions, processes, and more. This “unmanaged” model creates major blind spots, making daily operations, observability, debugging, and incident response challenging. When failures occur, limited visibility and bespoke systems make root-cause analysis slow, incomplete, or impossible.