Let's Talk Shop

Hosted by Elias Khnaser

Technology Business News

Website RSS feed

Episodes

Latest episode

Oct 2026

Language

EN-US

About the show

Coming to you from the Windy City, and hosted by Elias Khnaser! Welcome to Lets Talk Shop! A Podcast about all things cloud and enterprise tech. Listen to insights and guest interviews with IT thought leaders and professionals.

Listen to episodes

60 recent

November 5, 202642 min

How Google Cloud Is Rebuilding Its Network for the AI Era

AI gets all the headlines, but the network underneath it rarely gets the credit. In this episode of Let’s Talk Shop, I sit down with Kevin Hutchins, VP of Cloud Networking at Google Cloud, to talk about what it really takes to build infrastructure for AI at planet scale. We cover how Google Cloud’s network is evolving for massive training and inference workloads, why bandwidth and latency matter as much as the accelerators themselves, and how Google stress-tests its network with the traffic of Search, YouTube, Maps, Workspace and Gemini. Kevin explains why enterprises need to rethink network design for agentic AI, moving from predictable one-to-many apps to unpredictable many-to-many agent traffic. At the heart of the conversation is agent governance: knowing every agent on your network, giving each one a verified identity and policies, routing every interaction through an agent gateway built on Envoy, and getting full observability and semantic-level security. We also get into whether AI will split the cloud into “enterprise” and “AI” clouds, how model routing at the gateway can tame inference costs, and how operations teams will manage fleets of agents running at machine speed. If you’re an IT leader, architect or platform engineer trying to figure out where to focus in the AI era, Kevin’s message is clear: governance first. Chapters 00:00 Intro 01:42 Is the network AI’s most misunderstood component? 04:00 How Google Cloud’s network is evolving for AI 06:52 Has AI fundamentally changed network design? 09:51 Stress-testing a network at Google scale 13:25 Designing networks for AI vs. traditional enterprise apps 18:15 Security, identity and governance for AI agents 24:20 What sets Google Cloud apart 29:58 Will AI fracture the cloud? 34:12 Tokenomics and the cost of inference 38:41 Model routing at the agent gateway 39:34 Managing agent fleets at machine speed 42:00 Wrap-up Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

October 22, 2026Episode 428 min

Beyond GPUs: How Oracle Is Building Infrastructure for the AI Era

What does it really take to build AI infrastructure for the frontier-model era? In this episode of Let’s Talk Shop, Elias Khnaser sits down with Pradeep Vincent, Executive Vice President of Oracle Cloud Infrastructure (OCI), for a deep dive into how AI is fundamentally changing cloud infrastructure and the modern data center. The conversation goes far beyond simply adding more GPUs. Pradeep explains how Oracle is approaching the enormous networking, power, cooling, reliability and data-center challenges created by AI workloads — and why the infrastructure surrounding the GPUs can be just as critical to performance as the GPUs themselves. We discuss: ►Why AI infrastructure is fundamentally different from traditional cloud infrastructure ►Why GPUs alone don't determine AI performance ►The increasingly critical role of high-performance networking and RDMA ►How OCI designs infrastructure around the assumption that failures will happen ►Why AI training workloads require a different approach to network management ►How power consumption and massive AI workload swings are changing data-center design ►Liquid cooling and the challenge of supporting increasingly dense GPU racks ►How Oracle is designing data centers to remain viable for the next 15–20 years ►Whether AI infrastructure and traditional enterprise cloud infrastructure will eventually diverge ►Oracle's approach to water consumption, closed-loop cooling and responsible data-center expansion ►How hyperscalers can minimize the impact of massive AI campuses on local power grids and communities ►Why the modern AI data center is increasingly becoming one giant system — almost like the data center itself is the computer We also explore one of the biggest infrastructure questions facing the industry: How do you build today's AI data center without designing it around technology that could be obsolete tomorrow? If you're a CIO, CTO, cloud architect, infrastructure leader, AI engineer, data-center professional or simply interested in understanding the infrastructure powering the AI revolution, this conversation offers a fascinating look behind the scenes. Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

June 10, 2026Episode 329 min

Inside Raven Foresight: The Stealth Startup Building the First Cybersecurity Mesh Platform

Elias Khasner welcomes back Patrick Hevesi and Kevin Railsback to discuss Raven ForeSight — the flagship predictive cybersecurity platform from Cyber Mesh Security. In this conversation, Patrick Hevesi (Creator of Cybersecurity Mesh Architecture) and Kevin Railsback share their vision for how Raven Foresight is being built to deliver real-time, AI-powered threat intelligence as a core layer within a pure SaaS CSMA v3.1 platform. They discuss the early-stage architecture designed to ingest multiple threat feeds, apply advanced AI analysis with CSMA-aware scoring, and enable more automated policy and response actions across the cybersecurity mesh — helping organizations stay ahead of evolving threats like ransomware, supply chain attacks, and AI-driven campaigns. Raven ForeSight represents the next evolution in defense in depth: predictive, adaptive, and deeply integrated with the Cybersecurity Mesh Architecture. Built as a native SaaS CSMA platform Designed for multi-cloud environments Focused on reducing alert fatigue through intelligent analysis Leveraging multiple AI models and agentic capabilities Learn more about Raven ForeSight and Cyber Mesh Security: 🔗 https://CyberMeshSecurity.ai ►Patrick Hevesi (Creator of CSMA) LinkedIn: https://www.linkedin.com/in/patrickhevesi/ ►Kevin Railsback LinkedIn: https://www.linkedin.com/in/kevinrailsback/ Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

April 14, 2026Episode 244 min

Dark AI Exposed: The Hidden Cybersecurity Threat No One Is Talking About

Is AI the ultimate tool for humanity—or its greatest security threat? In this episode of Let’s Talk Shop, host Elias Khnaser sits down with cybersecurity expert Patrick Hevesi to explore the chilling reality of Dark AI. Patrick, a former Gartner colleague and the visionary behind Cybersecurity Mesh Architecture, reveals how easily "Dark AI" models are being built and weaponized on the dark web today. In this episode, we discuss: ► What is Dark AI? Why hackers are stripping away guardrails to create LLMs focused on chaos. ► The 15-Minute Cyberattack: Patrick explains how a standard $2,000 gaming rig can be used to run a boundary-less AI model. ► The "Chess Game" of Cyberwarfare: How rogue nations use AI to probe for vulnerabilities across major cloud providers like AWS, Google, and Microsoft. ► Patrick’s Next Move: A sneak peek into his transition from Gartner to his new stealth startup. Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

April 9, 2026Episode 153 min

The Future of DevOps: AI, Agents, and the "Impossible" Workflow

In this special episode, coming to you on the heels of Google Next, host Elias Khnaser sits down with an elite panel of industry experts to dive deep into the intersection of Generative AI and DevOps. We explore how AI is moving beyond simple code generation to orchestrating complex, "impossible" workflows. What We Discuss: ► The Rise of AI Agents: Moving from chatbots to autonomous agents that can manage entire software delivery lifecycles. ► Operationalizing AI: How to shift from "experimentation" to real-world remediation and automated incident response. ► The Changing Developer Experience: Will AI lead to "one-click" applications, and how should leaders prepare their teams for this shift? ► Trust and Hierarchy: Designing an "agent topology" where humans remain the final authority in an increasingly automated world. Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

December 31, 2025Episode 3955 min

AI Agents vs. Real Innovation: What Actually Happened at re:Invent 2025

Welcome to Let’s Talk Shop! In this episode, host Elias Khnaser is joined by cloud industry expert Sanjeev Mohan to break down the biggest takeaways from AWS re:Invent. The landscape of cloud and enterprise tech is shifting rapidly, and keeping up with the "onslaught" of new models and technologies can be deafening for IT professionals. We dive deep into why this year’s re:Invent felt different—from the "AI agent" saturated keynotes to the game-changing announcements that actually matter for your business. We discuss: ► The MultiCloud Pivot: Why AWS finally "threw in the towel" and embraced multi-cloud with the new AWS Interconnect. ► AI Factories: A look at AWS’s new approach to on-prem AI infrastructure. ► Nova Forge vs. RAG: Sanjeev explains why Nova Forge is a differentiator, allowing companies to build and tune proprietary frontier models for just $100k/year. ► The Future of Interconnectivity: How the industry is moving toward pointing compute at data, regardless of which cloud provider holds it. 2026 Predictions: What’s next for the market as AI continues to "take the oxygen out of the room". Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

December 18, 2025Episode 3818 min

How AWS is Moving Beyond 'Bolt-On AI' to Full Autonomy

How close are we to building Jarvis from Iron Man? Host Elias Khnaser sits down with Ali Maaz, AWS's leader for Go-to-Market Developer Services, to discuss how Amazon Q (Kiro) is transforming from a simple coding assistant to an Autonomous Agent and a peer on your team. Recorded live at AWS re:Invent 2025, this conversation dives deep into the future of enterprise software development and Cloud Operations. Key Takeaways: ► The Evolution of AI Agents: Why the biggest problem isn't coding, but the planning cycle—and how Kiro is solving it by arbitrating between product and engineering teams. ► Autonomous Agents: Our first look at Kiro Autonomous Agents, designed to address issues like bug fixes right out of Slack or Teams without a human opening a laptop. ► The Trust Factor: How AWS builds validation and trust with Property Based Testing, making Kiro a reliable, productive teammate. ► Cloud Ops Revolution: A major announcement focusing on a new agent specifically for Cloud Operations and DevOps to reduce Mean Time to Resolution (MTTR) and detect state/policy drift. ► AWS Differentiation: How AWS remains focused on customer-driven innovation, viewing internal teams (like Amazon.com) as just one of their largest and most important customers. ► The era of "bolt-on AI" is ending. The next step is "AI-driven development and operations." Tune in to see how you can "get out of the way" and let AI manage your next big project. 00:00:00 Intro & Guest Welcome: Ali Maz, AWS Developer Services 00:01:11 The "Jarvis" Question: How Close is AWS Q to Iron Man's AI? 00:01:46 Beyond Coding: Kiro's Role in the Planning Cycle (PR-FAQ) 00:03:43 Announcement 1: Kiro Autonomous Agent (From Assistant to Peer) 00:05:21 Building Trust: Validation, Oversight, and the Human in the Loop 00:06:46 Automated Reasoning & Property Based Testing (AI Validating AI) 00:07:34 Announcement 2: Kiro Powers & Personalized Context for ISVs 00:10:40 Agent Core: Policy Management & Evaluation for Production-Grade Agents 00:12:43 AWS Differentiation: Why We are Customer-Obsessed, Not Competitor-Obsessed 00:14:12 New Agent Announced: Focused on Cloud Operations & DevOps 00:16:33 The AI Evolution: Moving to "AI Managed" and "Get Out of the Way" 00:17:47 Conclusion & Wrap-up Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

December 18, 2025Episode 3735 min

Storage Architect to AI: Is Your Data Performance Fast Enough?

Is storage is the new bottleneck in the age of AI? Elias Khnaser and Asad Khan, Senior Director of Google Cloud Storage, discuss this topic in depth. While all the spotlight is on fast, expensive GPUs and TPUs, Elias and Asad are back to basics. In the past, CPU was never the bottleneck; slow storage was. Today, AI training and inferencing workloads require feeding high-cost GPUs/TPUs with data at an unprecedented speed to prevent them from sitting idle and wasting millions of dollars. Key Takeaways: ► The shift: Why high-performance storage is now mission-critical for maximizing your ROI on massive GPU clusters. ► How Google Cloud is solving the data performance problem by moving beyond HDDs to intelligent SSD tiering. ► Deep dive into Google Cloud Storage solutions for AI, including Anywhere Cache and Rapid Store, designed to automatically handle caching, prefetching, and high-performance throughput across all zones without the customer having to worry about colocation. ► The importance of data APIs for researchers: object storage (GCS) vs. full POSIX compliance (Lustre). ► The truth: The best AI performance isn't just about the fastest chip—it's the correct configuration of GPUs, storage, and networking. 00:00:00 Intro & Guest Welcome: Asad Khan, Google Cloud Storage 00:01:19 GCS, Lustre, & the Full Google Cloud Storage Portfolio 00:02:00 Is Storage Dead? The GPU vs. Storage Conversation 00:03:12 The New AI Bottleneck: Why GPUs Sit Idle (Wasting Money) 00:06:39 From Cheap Scale to High-Performance Cloud Storage 00:08:22 The Two Dimensions of AI Storage: SSDs & APIs 00:10:37 Anywhere Cache: Automatic High-Performance Caching 00:13:15 How Storage Differs for AI Training vs. Inferencing 00:15:35 Rapid Store and Full POSIX Compliance with Lustre 00:18:26 The True Formula for AI Performance (It's Not Just the GPU) 00:20:39 Sony Honda Mobility Case Study: Lustre in Action 00:23:41 Traditional vs. AI Customers: Different Storage Priorities 00:27:07 The Future: Unlocking Insights from Unstructured Enterprise Data 00:33:40 Final Thoughts & Key Takeaways Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

December 13, 2025Episode 369 min

AI in the Enterprise: Real World Use Cases & Reskilling

The honeymoon phase of AI is over. Elias Khnaser discusses with Pankaj Kumar, Executive Partner at IBM Consulting, the practical realities of deploying Agentic AI in the enterprise—beyond simple chatbots. Recorded live from AWS re:Invent 2025, this episode answers the biggest question facing executives: How do we do AI? We dive into a real-world case study of a major gas utility company (powering Las Vegas) that is completely reimagining its workflow to address one of its biggest problems: high-bill customer calls. Discover how the solution moves far beyond automating the contact center by using a multi-pronged approach that analyzes customer usage, infrastructure data, and weather patterns. Key Takeaways: ► Why the "boring work"—data governance, cloud architecture, and security—is the mandatory foundation for successful enterprise AI deployment. ► The strategic, phased approach: How the utility customer first executed a full cloud migration (DC exec to AWS) and SAP RISE before bolting on Agentic AI. ► The technology stack: How they integrated AWS Bedrock, LangChain, and LangGraph to create a comprehensive, agile solution. ► The Job Question: An honest conversation about the impact of Agentic AI on jobs. Is it mass firing, or a necessary focus on workforce reskilling and filling hard-to-fill contact center roles? #IBMPartner Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com

October 23, 2025Episode 3655 min

From Cloud to On-Prem: Gemini, GPUs, and the AI Anywhere Vision

The AI Revolution is here, but what about enterprises dealing with sensitive data, regulatory compliance, and low latency requirements? They can't always move to the public cloud—but now, they don't have to choose between compliance and innovation. In this episode of Let's Talk Shop, host Elias Khnaser sits down with two technology giants: Justin Boitano, VP of Enterprise AI at NVIDIA, and Rohan Grover, Senior Director and Head of Product for Google Distributed Cloud. They break down the deep technical partnership that is making the "AI Anywhere" vision a reality, allowing customers to run Google's cutting-edge Gemini 2.5 models directly on-premises using NVIDIA GPU servers. Discover how this collaboration uses confidential computing on both CPUs and NVIDIA Blackwell GPUs to secure sensitive customer data and proprietary model weights, turning previously inaccessible "dark data" into a source of competitive advantage. If you work in public sector, finance, healthcare, oil and gas, or any enterprise with strict data sovereignty rules, this discussion on on-prem GenAI and Google Distributed Cloud's managed and customer-owned deployment models is a must-watch. 👍 Like this video and Subscribe for more insights on cloud and enterprise tech! Sign Up Now for my online course "The Cloud Strategy Master Class": ► On my web site: https://lnkd.in/gcxcrX and use promo code LINKEDIN20 to receive a 20% discount. ► On Udemy: https://www.udemy.com/course/cloud-strategy-master-class/ PODCASTS: Listen wherever you get your podcasts: ► Let's Talk Shop: http://letstalkshop.buzzsprout.com ► Reality Distortion Fields RDFs Podcast: https://youtu.be/88z1UiVaV00 Follow me: ► TikTok: @ekhnaser ► Instagram: @ekhnaser ► Twitter: @ekhnaser ► LinkedIn: https://www.linkedin.com/in/eliaskhnaser/ ► Website: www.eliaskhnaser.com