Advertise on podcast: vBrownBag

Join us for Part 1 of a 3-part series as Du'An Lightfoot (Senior AI Engineer at Akamai) breaks down everything you need to know to get started running AI models locally on your own hardware. Du'An walks through the fundamentals of local AI - from understanding why you'd want to run models privately (data ownership, air-gapped environments, IP protection) to the hardware concepts that make it possible. You'll learn how inference actually works under the hood, why GPUs matter for AI workloads, how to choose and quantize models for your hardware, and how to get up and running with tools like Ollama. This is Part 1 of a 3-part series - future episodes cover serving models via API and distributing inference at the edge with Kubernetes. Timestamps 0:00 Cold Open: Why Local AI? 0:26 Welcome & Introduction 1:30 Following Up on the Frontier Models Episode 2:25 Du'An's Background & AI Inference at Akamai 3:49 What If You Wanted to Own Your Data? 5:00 Local AI vs Cloud AI: A Different Layer of the Stack 5:47 Why GPUs Matter: The Nvidia Story 7:03 CPU vs GPU: Serial vs Parallel Processing 8:28 Model Weights & Quantization Explained 12:45 Choosing the Right Model for Your Hardware 18:22 Getting Started with Ollama 24:16 Live Demo: Running Your First Local Model 30:41 Hardware Recommendations & Requirements 36:52 Hugging Face & Finding Models 42:18 Performance Tips & Benchmarking 48:35 Use Cases: When to Go Local vs Cloud 1:01:30 Live Demo: Claude Put-in-Work Repo 1:11:17 Bonus: Building a Deck with Co-work Live 1:16:36 Preview: Episodes 2 & 3 1:17:17 Wrap-up How to find Du'An: https://www.duanlightfoot.com/ https://github.com/labeveryday/ Links from the show: https://ollama.com/ https://apxml.com/ https://localllm.in/ https://huggingface.co/ https://github.com/labeveryday/claude-put-in-work https://claude.ai/

Join us as Kira Intrator (MIT-trained urban planner, systems thinker, and social impact technologist based in Geneva) makes the case that AI for Good isn't failing because of models - it's failing because of systems. Kira walks through why so many AI pilots never reach deployment, drawing on her experience building tools scaled across 9,000 users, three ministries, and six countries in Central Asia. You'll learn the five factors that kill AI projects in the development sector, why 80% of clinical AI models are trained on data that can't be deployed outside Western contexts, and what the $2.6 trillion opportunity in developing markets actually requires to unlock. This episode is equal parts systems thinking masterclass and call to action - a rare perspective from someone who has moved AI from prototype to production in places most tech professionals never consider. Timestamps 0:00 Welcome & Introduction 2:47 Kira's Background: MIT, Geneva, Central Asia 3:54 The Core Thesis: It's About Systems, Not Models 5:20 AI is Our Generation's Revolution 6:35 The $2.6 Trillion Opportunity 7:17 The 80% Western Data Problem 8:20 Why AI Projects Fail in Development: 5 Factors 9:28 Systems Mismatch & Low-Bandwidth Environments 9:52 Built for Pilot vs. Built for Deployment 10:29 Ownership, Economics & Sustainability 18:22 Real-World Case Studies 24:16 What Actually Works: Levers for Scale 30:41 The Role of Tech Companies & Foundations 33:39 Crystal Ball: Merging the Two Universes 35:01 A Call to Action 38:48 Wrap-up How to find Kira: https://www.linkedin.com/in/kiraintrator/ Links from the show: Infrastructure & Platforms Anthropic Beneficial Deployments: https://www.anthropic.com/ Google Research Global South Labs: https://research.google/ Lelapa AI: https://lelapa.ai/ Microsoft AI for Good: https://www.microsoft.com/en-us/ai/ai-for-good OpenAI Foundation: https://openai.com/ Research & Innovation Hubs Data Science Africa: https://www.datascienceafrica.org/ Masakhane: https://www.masakhane.io/ Stanford HAI: https://hai.stanford.edu/ Wadhwani AI: https://www.wadhwaniai.org/ Global Governance & Policy OECD AI Observatory: https://oecd.ai/ UNICEF Office of Innovation: https://www.unicef.org/innovation/ World Health Organization AI: https://www.who.int/ Funders & Philanthropies Gates Foundation: https://www.gatesfoundation.org/ Patrick J. McGovern Foundation: https://www.mcgovern.org/ Conferences AI for Good Global Summit (July 7-10, 2026 - Geneva): https://aiforgood.itu.int/ Data Science Africa 2026 (July 20-24 - Kampala, Uganda): https://www.datascienceafrica.org/ Deep Learning Indaba 2026 (August 2-7 - Lagos, Nigeria): https://deeplearningindaba.com/

Join us as Hiroko Nishimura (AWS Hero, LinkedIn Learning Instructor, and author of AWS for Non-Engineers) reflects on seven years of teaching cloud to 700,000 learners - what she's learned about learning, and how AWS education has changed. Hiroko walks through the evolution of AWS certification content, what changed when the Cloud Practitioner exam shifted focus, and her honest take on the industry's move away from non-engineer focused learning. You'll hear her best advice for anyone wanting to build a career in tech through content creation, why you only need to be 1-2 steps ahead to start teaching others, and how community has shaped her entire journey. This episode is equal parts AWS education deep dive and career inspiration - whether you're studying for your first cert or wondering how to break into the cloud community, Hiroko's 700,000-learner perspective is exactly what you need to hear. Timestamps 0:00 Welcome & Introduction 1:42 How We Get to 700,000 Learners 3:37 The 22 Courses Explained 6:15 How AWS Cloud Practitioner Has Changed 7:31 The Shift Away from Non-Engineer Focus 12:45 What Actually Changed in the Exam Content 18:22 Hiroko's Teaching Philosophy 24:16 How AI Has Changed the Learning Landscape 30:41 Community Building & AWS Heroes 35:00 Content Creation as a Career Strategy 39:02 Key Takeaways: You Only Need to Be 1-2 Steps Ahead 41:08 The Origin Story: 700,000 Learners from One Study Blog 44:02 Wrap-up & Where to Find Hiroko How to find Hiroko: https://www.linkedin.com/in/hirokonishimura/ https://hirokonishimura.com/ https://hiroko.io/ Links from the show:

Join us as Josh and Adriana call BS on the oversimplified vendor neutrality narrative - because switching observability vendors isn't magic, even with OpenTelemetry. Josh and Adriana walk through the hard truths about OTel vendor neutrality using their favorite analogy: switching from iOS to Android because vCard exists. Sure, your contacts will move, but what about everything else? You'll learn what vendor neutrality actually means in production, why vendor-neutral instrumentation still matters (your code artifacts survive tool changes), the real challenges and pitfalls of switching vendors, and best practices to make the process as pain-free as possible. This episode cuts through the hype with honest talk about what works, what doesn't, and why OpenTelemetry is still valuable even when it's not a magic wand. Timestamps 0:00 Welcome & Introduction 3:32 Getting Into the Talk 4:28 Origin Story: From LinkedIn Post to Full Presentation 5:35 Standards We Love: USB-C, Stop Signs, McDonald's 8:00 The No Name Brand Analogy 12:45 What Vendor Neutrality Actually Means 18:22 The iOS to Android / vCard Comparison 24:16 What You Lose When Switching Vendors 30:41 Why Vendor-Neutral Instrumentation Matters 36:52 Real Challenges & Pitfalls 42:18 Best Practices for Switching 46:17 Shameless Self-Promotion & Resources 49:03 Wrap-up How to find Josh & Adriana: https://www.linkedin.com/in/joshuamlee/ https://www.linkedin.com/in/adrianavillela/ Links from the show: https://opentelemetry.io/

Join us as Amelia shares the debugging story nobody tells you about - how her vector store DB couldn't surface specific data until she tested it with simplified data from ChatGPT. Amelia walks through her journey from throwing JIRA tickets into a large language model without understanding pipelines or data cleaning, to discovering why her production vector store was failing. You'll learn about the gap between chatting with data and getting accurate connections, how to validate vector similarity search results, the difference between production and synthetic test data, and practical troubleshooting workflows for AWS vector stores. This episode reveals the messy reality of RAG systems - when everything seems fine but the outputs are subtly wrong, and how testing with simplified data can expose what production complexity hides. Timestamps 0:00 Cold Open 1:03 Welcome & Introduction 2:06 Amelia's Background & DeepRacer Trophy 4:49 The JIRA Ticket Use Case Origin Story 5:53 Getting Into the Presentation 6:03 Accessing & Cleaning Data Sets 8:12 Losing Production Data & Recreating with ChatGPT 12:45 Understanding Vector Databases 18:22 How Embeddings Work 24:16 The Hallucination Discovery 30:41 Testing Strategies for Vector Stores 36:52 Debugging Vector Similarity Search 42:18 Real-World Troubleshooting Workflows 44:26 Where to Find Amelia & Wrap-up How to find Amelia: https://www.linkedin.com/in/ameliahoughross/

Join us as Du'An breaks down AI agents in a way that actually makes sense - what they are, how to use them, and how to get started today. Du'An walks through the fundamentals of AI agents with live demos and practical code examples you can use immediately. You'll learn about agent frameworks, when to use agents versus simple LLM calls, building your first agent, and real-world applications from bookmark management to automated workflows. This episode cuts through the hype with realistic expectations about what agents can and can't do, while showing you concrete examples including MCP servers, Strands Pack, and Du'An's personal second brain system. Timestamps 0:00 Welcome & Introduction 1:39 Du'An's Background & Previous Episode Success 3:06 Segueing from Last Week's Episode 4:03 CEOs Vibe Coding Discussion 6:49 Real Estate Developer Building Apps Story 8:23 Getting Started with the Presentation 12:45 What Are AI Agents? 18:22 Agent Frameworks Overview 24:16 When to Use Agents vs Simple LLM Calls 30:41 Building Your First Agent 36:52 Live Demo: Strands Pack 42:18 MCP Servers Explained 47:35 WriteStats MCP Demo 52:14 Real-World Applications 58:33 Du'An's Second Brain System 1:04:01 Bookmark Manager Walkthrough 1:07:17 Organizing Cloud Storage & Email 1:09:06 Wrap-up & Next Episode Teaser How to find Du'An: https://www.duanlightfoot.com/ https://github.com/labeveryday/ Links from the show: https://github.com/labeveryday/strands-pack https://github.com/labeveryday/writestat-mcp https://github.com/labeveryday/bookmark-manager-site https://bookmarks.duanlightfoot.com/ https://github.com/openai/whisper https://openai.com/index/whisper/

Join us as Chris gets brutally honest about tech employment in the AI era: what's dying, what's thriving, and how to position yourself to survive the chaos. Chris walks through the current state of tech layoffs hitting record numbers while companies post record profits, the disappearance of entry-level roles, and practical strategies for navigating this unprecedented moment. You'll learn about skill development in the AI era, why fundamentals still matter more than hype, how to build resilience through community, and what hiring managers are actually looking for right now. This episode doesn't sugarcoat the challenges from hollowed-out expertise at major companies to early-career professionals wondering if their degree still matters, but it also provides actionable guidance on positioning yourself and why humor and human connection remain irreplaceable in an AI-driven world. Timestamps 0:00 Welcome & Setting the Tone 3:09 Chris Miller's Background & Journey 7:30 The Current State of Tech Employment 12:45 Layoffs vs Record Profits Discussion 18:22 Entry-Level Roles Disappearing 24:16 What Skills Actually Matter Now 30:41 Building Career Resilience 36:52 The Fundamentals Still Win 42:18 Community & Support Networks 47:35 Practical Job Search Strategies 52:14 What AI Can't Replace (Yet) 55:06 Things We're Thankful For 59:00 Wrap-up & Resources How to find Chris: https://www.linkedin.com/in/chris-t-miller/ https://www.chrismiller.com/ Links from the show: https://roadmap.sh

Join us as Marian explains what AI governance means for vSphere administrators and why it matters now. Marian walks through practical governance frameworks that vSphere admins need to understand, from IEEE 7000 series standards to mapping governance controls onto infrastructure you already manage. You'll learn what your CISO will ask for, how to respond using your existing VMware stack, and why governance isn't about slowing innovation - it's about enabling it safely. This episode covers real-world scenarios from data lineage and model transparency to integrating governance tools with existing infrastructure, and addresses the gap between compliance requirements and practical implementation for virtualized environments. Timestamps 0:00 Welcome & Introduction 5:16 Marian's Background in Tech & Governance 6:37 What is Governance? 12:45 IEEE 7000 Series Standards Overview 18:22 AI Governance for vSphere Admins 24:16 Data Lineage & Model Transparency 30:41 Risk Assessment Frameworks 36:52 Practical Implementation Strategies 42:18 Integration with Existing Tools 47:35 Common Governance Challenges 51:12 Vendor Landscape Discussion 54:27 Missing Innovation in the Space 58:09 Wrap-up & Resources How to find Marian: https://www.linkedin.com/in/mariannewsome/ Links from the show: https://ethicaltechmatters.com/

Join us as Peter explores the core principles and practices of FinOps that help organizations optimize cloud spend without slowing innovation. Peter walks through what FinOps really is, why it matters beyond just cost cutting, and how engineers can collaborate effectively with finance teams to design cost-aware architectures. You'll learn about the three phases of FinOps (Inform, Optimize, Operate), how to get leadership buy-in for cloud initiatives, and practical strategies for managing cloud costs from the architecture phase through operations. This episode covers real-world scenarios from hybrid cloud cost tracking to building cost models before migrations, and explains how FinOps fits into your existing team structure regardless of organization size. Timestamps 0:00 Welcome & Introduction 6:10 Peter's Background & Journey to FinOps 10:45 What is FinOps? 16:32 The Three Phases: Inform, Optimize, Operate 22:18 Getting Leadership Buy-In 28:45 Cost-Aware Architecture Design 34:20 Hybrid Cloud & On-Prem Cost Tracking 40:15 FinOps Team Structure & Roles 46:30 Tools & Platforms Discussion 52:14 Accounting & Finance Collaboration 54:13 Starting FinOps Before Cloud Migration 57:17 FinOps for Small Teams & DBAs 1:00:13 Wrap-up & Resources How to find Peter: https://www.linkedin.com/in/petercrenshaw/ Links from the show: https://finops.org https://finopsweekly.com https://thefrugalarchitect.com

Join us as Neel explores how observability is evolving beyond traditional logs, metrics, and traces into a predictive, AI-powered discipline. Neel walks through the evolution of Observability, demonstrating how OpenTelemetry, machine learning, and LLMs are transforming how we monitor and maintain modern applications. You'll learn about dynamic sampling techniques that reduce costs while maintaining visibility, how ML algorithms detect anomalies before they cause outages, and practical implementations using tools like the OpenTelemetry Collector. This episode covers real-world scenarios from reducing massive log volumes to predicting system failures before they impact customers. Timestamps 0:00 Welcome & Introduction 4:29 Neel's Background & Community Work 5:03 The Evolution of Observability 6:29 The 2 AM Production Incident Scenario 8:13 OpenTelemetry's Role in Modern Observability 12:45 Dynamic Sampling Techniques 18:22 ML & AI in Anomaly Detection 24:16 LLM Observability Explained 28:32 Cost Optimization Strategies 30:04 Context Windows & Token Management 32:00 Self-Healing Systems Discussion 34:15 Edge Cases: When Dynamic Sampling Doesn't Work 36:27 Wrap-up & Resources How to find Neel: https://www.linkedin.com/in/neelcshah/ https://bento.me/neelshah Links from the show: https://neelshah.dev/blogs/observability-2 https://opentelemetry.io/ https://middleware.io/blog/observability-2-0/

Join us as Sam demonstrates how to teach AI to write Terraform configurations using Model Context Protocol (MCP) servers. Sam introduces the Terraform MCP server and walks through practical demos showing how AI can understand and safely interact with your infrastructure. You'll see live examples of AI planning, generating, and evolving Terraform configurations - from creating landing zones to setting up workspace variables automatically. Whether you're managing complex multi-cloud environments or just getting started with infrastructure as code, this episode demonstrates how MCP servers bridge the gap between AI capabilities and real-world Terraform workflows. Learn how to get started, which Claude models work best for different tasks, and best practices for integrating AI into your IaC pipelines. Timestamps 0:00 Welcome & Introduction 4:37 Sam McGeown's Background 6:02 Introduction to Terraform MCP Server 12:35 What is Model Context Protocol? 18:22 Setting Up the Terraform MCP Server 24:16 Demo: Claude Desktop Integration 30:41 Creating Infrastructure with AI Prompts 36:52 Reading & Analyzing Existing Terraform Code 42:18 Generating Landing Zone Configurations 47:35 Working with Terraform Workspaces 50:37 Creating Variables Automatically 52:14 Model Selection: Sonnet vs Opus 55:11 Live Demo: Workspace Variable Creation 58:33 Getting Started & Resources How to find Sam: https://www.linkedin.com/in/sammcgeown/ Links from the show: https://developer.hashicorp.com/terraform/mcp-server

Join us as Gautam breaks down the evolution of tool use in generative AI and dives deep into MCP. Gautam walks through the progression from simple prompt engineering to function calling, structured outputs, and now MCP—explaining why MCP matters and how it's changing the way AI systems interact with external tools and data. You'll learn about the differences between MCP and traditional API integrations, how to build your first MCP server, best practices for implementation, and where the ecosystem is heading. Whether you're building AI-powered applications, integrating AI into your infrastructure workflows, or just trying to keep up with the latest developments, this episode provides the practical knowledge you need. Gautam also shares real-world examples and discusses the competitive landscape between various AI workflow approaches. Subscribe to vBrownBag for weekly tech education covering AI, cloud, DevOps, and more! ⸻ Timestamps 0:00 Introduction & Welcome 7:28 Gautam's Background & Journey to AI Product Management 12:45 The Evolution of Tool Use in AI 18:32 What is Model Context Protocol (MCP)? 24:16 MCP vs Traditional API Integrations 30:41 Building Your First MCP Server 36:52 MCP Server Discovery & Architecture 42:18 Real-World Use Cases & Examples 47:35 Best Practices & Implementation Tips 51:12 The Competitive Landscape: Skills, Extensions, & More 52:14 Q&A: AI Agents & Infrastructure Predictions 55:09 Closing & Giveaway How to find Gautam: https://gautambaghel.com/ https://www.linkedin.com/in/gautambaghel/ Links from the show: https://www.hashicorp.com/en/blog/build-secure-ai-driven-workflows-with-new-terraform-and-vault-mcp-servers Presentation from HashiConf: https://youtu.be/eamE18_WrW0?si=9AJ9HUBOy7-HlQOK Kiro Powers: https://www.hashicorp.com/en/blog/hashicorp-is-a-kiro-powers-launch-partner Slides: https://docs.google.com/presentation/d/11dZZUO2w7ObjwYtf1At4WnL-ZPW1QyaWnNjzSQKQEe0/edit?usp=sharing

Join the vBrownBag crew and AWS Hero Peter Sankauskas for the 2026 AWS Community Survey 🎯 – your chance to shape the conversation around AWS careers, cloud adoption trends, and emerging technologies. Peter breaks down this year's survey methodology, shares surprising insights from previous years, and explains how your responses help the entire AWS community understand compensation trends, service adoption patterns, and career development paths. Whether you're a cloud architect, DevOps engineer, or just starting your AWS journey, your voice matters ☁️ Take the survey from January 6 to February 6, 2026, and help build the most comprehensive snapshot of the AWS ecosystem! Chapters: 00:00:03 - Welcome and 2026 AWS Survey Introduction 00:04:14 - Survey Design and Methodology Explained 00:12:45 - Key Insights from Previous AWS Surveys 00:21:30 - New Questions for 2026: AI, FinOps, and Career Trends 00:28:15 - How to Complete the Survey Effectively 00:35:03 - Closing Thoughts and Survey Timeline Resources: Take the survey! https://answersforaws.com/survey/ https://www.linkedin.com/in/petersankauskas/ Boardgames mentioned: https://boardgamegeek.com/boardgame/201808/clank-a-deck-building-adventure https://boardgamegeek.com/boardgame/148228/splendor https://boardgamegeek.com/boardgame/264806/were-doomed #AWS #CloudComputing #DevOps #CloudCareers #AWSCommunity #vBrownBag #AWSSurvey

Join us for the final episode of 2025 as Mark Tinderholt (Principal Software Engineer at Microsoft Azure, HashiCorp Ambassador, and author of "Mastering Terraform") teaches us Infrastructure as Code through Minecraft! If you've ever wanted to learn Terraform in a fun, visual way, this is the episode for you. Mark demonstrates how to use the Minecraft Terraform provider to build infrastructure in-game, making complex IaC concepts tangible and engaging. You'll see live demos of provisioning Minecraft resources, managing dependencies, handling state, and even importing existing structures into Terraform. This unique approach transforms abstract infrastructure concepts into something you can literally see and interact with—perfect for visual learners, educators, or anyone looking to make IaC training more engaging. Whether you're teaching your team Terraform or just want a creative way to understand infrastructure patterns, this episode shows you how gaming and cloud engineering can come together. Subscribe to vBrownBag for weekly tech education! ⸻ Timestamps 0:00 Welcome & Technical Difficulties 1:27 Last Episode of 2025! 4:41 Planning for 2026 5:37 Mark Tinderholt Joins 6:14 Introduction to Minecraft + Terraform 8:52 Why Use Minecraft for Teaching IaC? 12:35 Getting Started: Requirements & Setup 16:47 The Minecraft Terraform Provider 20:18 First Demo: Provisioning Basic Blocks 28:32 Managing State in Minecraft 35:41 Working with Dependencies 42:16 Advanced Patterns: For_each & Count 48:55 Importing Existing Structures 55:23 Real-World Applications & Teaching 1:00:17 Q&A: Provider Limitations & Features 1:05:24 Minecraft Level Building Tools Discussion 1:09:05 Final Giveaway & Wrap-Up How to find Mark: https://www.linkedin.com/in/marktinderholt/ Links from the show: Marks repos: https://github.com/markti?tab=repositories Marks book: https://amzn.to/3N1rnuJ Mark's Ignite talk: https://ignite.microsoft.com/en-US/sessions/7fa5095f-9f65-46e3-9f82-9af6603ea903

Join Du'An Lightfoot, AI Developer at AWS, as he dives deep into building AI agents with the strands framework. In this technical walkthrough, Du'An demonstrates how to create custom AI coding assistants and multi-agent systems in just a few lines of code. Learn how agentic AI frameworks have evolved from basic function calling to sophisticated systems that can rival tools like Cursor and Cloud Code. Du'An shares practical examples, including building content pipelines, preprocessing systems, and even generating a book outline from his own YouTube content. Whether you're looking to automate workflows or build your own AI-powered tools, this session covers the frameworks and techniques you need to get started with AI agents. Perfect for developers, DevOps engineers, and anyone interested in leveraging AI to enhance their development workflow. Subscribe to vBrownBag for more community-driven tech education! ⸻ Timestamps 0:00 - Introduction & Welcome 6:43 - AI Tools Discussion & Current Usage 9:33 - Technical Background & Getting Started with Agents 15:00 - Introduction to Strands Framework 25:00 - Building Custom AI Agents Demo 40:00 - Multi-Agent Systems & Workflows 55:00 - Content Pipeline & Preprocessing Examples 1:05:00 - Book Generation Demo 1:10:00 - Q&A & Wrap Up How to find Du'An: https://www.linkedin.com/in/duanlightfoot/ Links from the show: https://s12d.com/vbrownbag-2025 https://github.com/strands-agents/samples https://modelcontextprotocol.io/docs/getting-started/intro https://www.anthropic.com/engineering/building-effective-agents https://github.com/awslabs/amazon-bedrock-agentcore-samples https://github.com/modelcontextprotocol/python-sdk https://modelcontextprotocol.io/llms-full.txt https://openai.com/index/whisper/ https://github.com/openai/whisper