GovAi.fm

AI isn't coming to BC — it's already here. I sit down with the Honourable Rick Glumac, BC's first Minister of State for AI and New Technologies, to talk about what that means and where it's headed.We get into how BC is supporting AI adoption across sectors — through programs like the Integrated Marketplace, work in healthcare, and Connected Services BC. We talk about digital sovereignty, what it takes to build AI infrastructure on clean energy, and how the current geopolitical moment is creating new opportunities for the province.We also tackle the question everyone's asking: what does AI mean for my job? Rick makes a compelling case that human creativity and ingenuity remain at the centre of it all — and that AI is a tool, not a replacement.Whether you work in government, enterprise, or tech — this one's worth your time.🎙️ Guest: Honourable Rick Glumac🎙️ Host: Girish Limaye

Traditional software development is linear, slow, and heavily reliant on manual coding. What happens when AI agents take the wheel?Host Girish Limaye sits down with Andre Kaminski to unpack the realities of AI-Native Software Development. Moving past the hype of simple coding assistants like GitHub Copilot, Andre breaks down the exact framework required to build and manage truly AI-native organizations.From managing token budgets and multi-agent orchestration to surviving the developer "identity crisis," this episode is a tactical roadmap for tech leaders and engineers navigating the next massive shift in tech.What you'll learn:-The 4 foundational shifts: from imperative coding to prompt-driven orchestration-How to mitigate LLM hallucinations using hierarchical prompt architecture and test-driven navigation-The 6-phase AI-Native SDLC every tech leader needs to implement-The "U-Shaped" developer: How to future-proof your engineering career-How to build AI governance with a 4-tier risk-based automation model-Why "token budgets" will replace infrastructure costs🎙️ Guest: Andre Kaminski🎙️ Host: Girish Limaye

Most AI deployments have a blind spot. This episode is about finding yours.Host Girish Limaye sits down with Dr. Vered Shwartz, UBC Assistant Professor and Canada CIFAR AI Chair, for a frank, research-grounded conversation about the gap between AI hype and reality.From cultural blind spots (CulturalBench) to hallucinations and high-stakes AI in healthcare and regulatory contexts, this is what every leader needs to hear before their next AI deployment.What you'll learn:- Where AI bias actually comes from- Why models fail on cultural knowledge- The honest state of hallucination research- How to evaluate AI before deploying it in your organization- What AI leaders need to know in 2026🎙️ Guest: Dr. Vered Shwartz, UBC / Canada CIFAR AI Chair🎙️ Host: Girish Limaye, GovAI.fm

AI agents are evolving from simple task runners to autonomous systems that plan, act, and adapt. In this episode, we explore where they are today, where they’re headed, and what public and private sector leaders should be thinking about when it comes to transparency, control, and real-world use.

Everywhere you look, teams are chasing reliability in GenAI.The promised solution: RAG — Retrieval-Augmented Generation.Ground the model in your company’s data, and it’ll stop hallucinating… right?Not quite.As Hubert Duan, Principal at Microsoft AI, and I discuss — RAG introduces its own complexities: retrieval quality, data freshness, latency, and failure modes that only surface at scale.In this GovAI.fm conversation, we go beyond the hype to explore what it really takes to make RAG work in production:🔍 What RAG really is (and how it’s often misunderstood)⚠️ The list of failure modes that teams encounter when scaling RAG🏗 How to design resilient, scalable RAG architectures🚀 Why RAG alone won’t fix reliability — and what’s next, from hard scaling lessons to Agent-to-Agent (A2A) intelligenceWe also talk about Hubert’s public sector successes, including projects like AI for wildfire forecasting in Alberta — and how building AI for good causes can help attract and retain talented public sector AI professionals.Finally, Hubert shares his perspective on the future of AI, from RAG’s evolution to the emerging world of autonomous, agent-to-agent systems.If you’ve ever wondered why your GenAI prototype struggles in production — or what’s coming next — this episode connects the dots.

If Part 1 set the stage (60 years · 5 tribes), Part 2 zooms in on the “connectionists” — the neural-network tribe that quietly persisted through AI winters and then reshaped the world.In this episode, Aman Sidhu (senior public-sector leader) and  Girish Limaye (AI consultant & educator), describe deep learning and its recent emergence in a story format — so it’s accessible to everyone and helps you appreciate the key moments that shaped modern AI.In this episode, we explore:✅ From single “neurons” to deep nets: the simple idea + backprop intuition✅ Why deep learning conquered unstructured data (vision, speech, text)✅ The ImageNet moment (2012) and what a 60% drop in error really meant✅ Transfer learning: reusing layers to go from potholes → checks → cancer scans✅ Accuracy vs. explainability: when “black box” blocks adoption—and how guardrails help✅ What this means for government: from asset inspection to digital ID verification—and why data quality + governance still decide winners2 quick takeaways:-“It can learn anything” ≠ “It will learn the right thing"- Foundation models shift AI from project-by-project builds to platform 📺 Missed Part 1? Start here: https://youtu.be/cI106r0XSbA🔔 Follow us for latest podcast updates— we alternate explainers, real use cases, and current trends for public-sector leaders.Join the conversation:• What’s one workflow you wouldn’t trust to a black box today?• Where could transfer learning save you months?

👉 Is AI just another productivity tool—or a complete disruptor of how we build? In this episode, Andre Kaminski and Girish Limaye unpack how AI is reshaping principles, teams, and the future of technology.What does building software look like in the age of AI? In this episode of GovAI.fm, Girish Limaye sits down with Andre Kaminski to explore how AI is reshaping teams, tools, and the very principles of technology.We dive into:Timeless Tech Principles that still hold true todayAI: Productivity Tool or Disruptor?Vibe Coding and new ways of creatingThe shift from DevOps → NoOpsGovernance in the Age of AIAI-Powered Testing and quality assuranceThe Future of Teams and collaboration modelsEssential Skills for the AI EraIf you’re curious about how AI is changing not just coding, but the culture of building in government and beyond, this episode is for you.🔔 Follow GovAI.fm for more conversations with innovators shaping the future of AI in the public sector.

💡 Innovation isn’t about ideas—it’s about impact.Whether you’re launching something new, scaling a small program, or pushing change without a formal team, success comes from mastering the full life cycle: sparking ideas, choosing the right model, measuring what matters, and keeping momentum alive.In this GovAI.fm episode, Dr. Nan Xie shares lessons from leading award-winning smart city projects and applying Industrial IoT & AI to real-world challenges—plus the pitfalls to avoid.Topics in this episode:• Why Innovation Matters• Picking the Right Model• Measuring What Matters• Building an Innovation Culture• Lessons from Failure• Building the Team• Emerging Tech Outlook• Recommended Resources

Sixty years. Five tribes. Countless breakthroughs and setbacks.If you want an intuition for where AI is headed, you first need to know where it came from.In Part 1 of The Story of AI, we uncover:✅ The five “tribes” that shaped modern AI✅ The surprising patterns hidden in decades of AI booms and busts✅ How yesterday’s ideas quietly power today’s breakthroughs✅ What history can teach us about building responsible, practical AI in government🎯 Why watch?Because the future of AI makes a lot more sense when you know its past.

We kicked things off with a conversation between Aman Sidhu, senior public sector leader and Girish Limaye AI Consultant ✅ Why AI still feels confusing—and how to fix that✅ Our personal journeys into AI (from early data days to deep learning revolutions)✅ What public sector teams get right—and wrong—when starting with AI✅ The importance of data foundations, human-in-the-loop design, and moving beyond cool demos to real value✅ Why applied AI doesn’t need to be complicated—and how we’re here to make it “boring again” (in a good way!)🎯 Our goal?To demystify AI and make it approachable, responsible, and practical for public sector innovators. Whether you're leading digital transformation or just getting started with pilots, we hope this show becomes your trusted companion in the AI journey.💬 Join the conversation:What’s one thing you wish more people in government understood about AI? Drop it in the comments

The Trust Layer for AI Projectswith Dr. Curtis Northcutt, CEO & Co-founder of Cleanlab on the GovAI.fm podcastWhat if your AI could flag hallucinations before they launch?What if the same tech that finds hidden errors in benchmark datasets like MNIST can now catch LLM hallucinations, bias, and uncertainty—at scale?In this conversation, we explore:✅ His journey from rural Kentucky to a PhD at MIT, to founding Cleanlab, raising a $25M Series A, and powering trust scores used by 100+ Fortune 500 companies✅ Why AI systems quietly fail when benchmarks are broken✅ The real story behind Cleanlab’s origins — from capsule networks to confident learning✅ What most public-sector teams miss when deploying LLMs✅ How trust scores could become the compliance layer before regulation lands—🎧 This episode is part of GovAI.fm — your trusted source for real, practical AI in the public sector.We bring you behind-the-scenes with leaders building AI that works for citizens.