Inspire AI: Transforming RVA Through Technology and Automation

Ep 78 - Learn By Building: From Strategy Decks To Working Agents w/ Matt Bartles

44 min · 11 de may de 2026
Portada del episodio Ep 78 - Learn By Building: From Strategy Decks To Working Agents w/ Matt Bartles

Descripción

Send us Fan Mail [https://www.buzzsprout.com/2433144/fan_mail/new] “We’ll learn AI once we understand it” sounds responsible, but it’s one of the fastest ways to fall behind. We sit down with Matt to argue for a different approach: learn AI by building with it, in small scopes, with real users, and with the humility to let the work teach you what the strategy can’t. The result is faster AI adoption, better judgment about what models can and cannot do, and a team that develops true operational muscle instead of slide-deck confidence. We dig into why long AI roadmaps are so fragile, how experimentation creates better plans, and what the real costs look like when you delay hands-on work. That includes the unglamorous details that decide whether an AI feature scales: token costs, context loading, caching, latency, and picking the right model for the job. We also explore when open models make sense, what it takes to host them, and why workflow design matters just as much as model choice in complex environments like banking and underwriting. Then we get practical about building agents. A simple “meal planner” agent becomes a lesson in inconsistency, unclear pathways, and why agents can fall apart when they must choose from a long list of similar options. From there, we talk guardrails: where probabilistic AI is fine, where deterministic rules must take over, and how governance should tighten as usage grows. If you’re leading teams through AI strategy, enterprise AI, or agentic AI pilots, you’ll leave with a clearer playbook for building safely and learning fast. Subscribe for more conversations like this, share the episode with a teammate, and leave a review if it helps. What’s the smallest AI build you could ship in the next 30 days? Want to join a community of AI learners and enthusiasts? AI Ready RVA [https://aireadyrva.com/] is leading the conversation and is rapidly rising as a hub for AI in the Richmond Region. Become a member [https://aireadyrva.com/membership-options/] and support our AI literacy initiatives.

Comentarios

0

Sé la primera persona en comentar

¡Regístrate ahora y únete a la comunidad de Inspire AI: Transforming RVA Through Technology and Automation!

Prueba gratis

Empieza 7 días de prueba

$99 / mes después de la prueba. · Cancela cuando quieras.

  • Podcasts solo en Podimo
  • 20 horas de audiolibros al mes
  • Podcast gratuitos

Todos los episodios

86 episodios

episode Ep 84 - The Philosophical Shift: As Intelligence Becomes Cheap, Evaluation Becomes Everything artwork

Ep 84 - The Philosophical Shift: As Intelligence Becomes Cheap, Evaluation Becomes Everything

Send us Fan Mail [https://www.buzzsprout.com/2433144/fan_mail/new] AI can generate code, analysis, and recommendations faster than any team in history, but there’s a catch: verification doesn’t scale the same way. When intelligence becomes abundant, judgment becomes scarce, and that scarcity reshapes what “good engineering” and “good leadership” actually mean. We walk through the hidden asymmetry behind modern generative AI: organizations can produce far more software, content, and automated decisions than they can evaluate for correctness, safety, ethics, and alignment. That’s why AI evaluation is becoming infrastructure, not a side task. We dig into what trustworthy AI looks like in practice, including governance, observability, benchmark design, hallucination detection, adversarial testing, red teaming, and human review workflows that keep risk from silently compounding. Then we zoom out from software engineering to leadership. Evaluation is an organizational question: who defines acceptable risk, who owns accountability, who sets escalation paths, and who decides when humans stay in the loop? As AI becomes operational infrastructure, leaders become stewards of intelligent systems, and the core advantage shifts from speed to trust. If you’re building with generative AI, take this as a blueprint for creating an evaluation culture that scales. Subscribe, share this with a builder or leader on your team, and leave a review with the biggest verification challenge you’re facing right now. Want to join a community of AI learners and enthusiasts? AI Ready RVA [https://aireadyrva.com/] is leading the conversation and is rapidly rising as a hub for AI in the Richmond Region. Become a member [https://aireadyrva.com/membership-options/] and support our AI literacy initiatives.

22 de jun de 202611 min
episode Ep 83 - Up the Stack: The Five Layers Of The Future Software Engineer artwork

Ep 83 - Up the Stack: The Five Layers Of The Future Software Engineer

Send us Fan Mail [https://www.buzzsprout.com/2433144/fan_mail/new] AI can write code faster than any team on earth, so why does it still feel like shipping software is hard? The uncomfortable answer is that speed is not the same as progress, and generation is not the same as judgment. We challenge the tired question “Will AI replace programmers?” and replace it with a more useful one: at what layer does human judgment become most valuable as AI absorbs more of implementation?  We introduce our “five layers of the future engineer” framework, starting with the builder and moving upward through the designer, architect, evaluator, and systems leader. Along the way, we unpack how AI coding tools compress implementation, why design quality and clear constraints start to decide outcomes, and how architecture becomes the guardrail that prevents cheap generation from turning into expensive chaos. If your organization is racing to add AI assistants, agents, and automated pipelines, this conversation gives you a practical way to think about roles, skills, and where leverage is actually shifting.  The biggest unlock is the evaluator mindset: verification, benchmarking, red teaming, AI governance, and hallucination detection become core engineering infrastructure as trust becomes a competitive advantage. We close by looking at systems leadership as orchestration, deciding where humans must own decisions, how accountability stays intact, and how to align velocity with mission in an AI-accelerated world. If this helped you rethink the future of software engineering, subscribe, share it with a teammate, and leave a review with the layer you are building next. Want to join a community of AI learners and enthusiasts? AI Ready RVA [https://aireadyrva.com/] is leading the conversation and is rapidly rising as a hub for AI in the Richmond Region. Become a member [https://aireadyrva.com/membership-options/] and support our AI literacy initiatives.

15 de jun de 202612 min
episode Ep 82 - A Devs Transformation: New Values Emerge As Code Becomes Cheap artwork

Ep 82 - A Devs Transformation: New Values Emerge As Code Becomes Cheap

Send us Fan Mail [https://www.buzzsprout.com/2433144/fan_mail/new] The scariest part of AI in software is not that it writes code. It is that it changes what “being an engineer” even means. When generative AI can scaffold applications, spin up infrastructure configs, draft tests, refactor modules, and debug common failures in minutes, the act of typing implementation stops being the limiting factor. That is a fundamental shift in software engineering, and it is bigger than automation. It is the next abstraction layer in the history of computing, and this time the tool participates in generation itself.  We zoom out and connect the pattern: every major leap in computing removes friction and pushes people to higher leverage work. As code generation becomes abundant, value migrates to judgment and coordination. The hard questions become: what problem matters, what should we build, how do systems interact, how do we evaluate AI outputs, how do we preserve trust, and how do we govern intelligent systems responsibly. The “future engineer” looks less like a manual builder and more like a systems designer, workflow orchestrator, and rigorous evaluator who can set constraints and verify outcomes.  We also talk about why this moment feels personal. Coding has been treated as the identity of engineering for decades, so AI codegen can feel existential. But abstraction rarely eliminates technical roles outright, it repositions them. And because software is embedded in healthcare, education, finance, logistics, government, and communications, faster software creation also accelerates organizational change and leadership demands. If you want to stay resilient in an AI accelerated world, the goal is not to be the fastest typer. It is to build clear thinking, strong evaluation habits, and the judgment that keeps systems trustworthy at speed.  If this sparked something for you, subscribe, share the episode with a builder or leader, and leave a review so more people can find the show. What skill do you think becomes most valuable when implementation is abundant? Want to join a community of AI learners and enthusiasts? AI Ready RVA [https://aireadyrva.com/] is leading the conversation and is rapidly rising as a hub for AI in the Richmond Region. Become a member [https://aireadyrva.com/membership-options/] and support our AI literacy initiatives.

8 de jun de 202610 min
episode Ep 81 - AI Needs Better Data: Agentic AI Foundations at Scale artwork

Ep 81 - AI Needs Better Data: Agentic AI Foundations at Scale

Send us Fan Mail [https://www.buzzsprout.com/2433144/fan_mail/new] AI agents are showing up everywhere, but most enterprises are discovering a frustrating truth: getting an agent to “work” in a demo is easy, getting it to deliver measurable value in production is brutally hard. We dig into why the bottleneck is shifting away from model performance and toward the fundamentals leaders control: data foundations, data governance, and organizational design that can support autonomous action. We break down what really changes when you move from generative AI to agentic AI. A chatbot that drafts copy is contained; an agent that updates CRM fields, coordinates inventory, triggers workflows, and pulls sensitive context has to operate inside your real enterprise systems. That’s where fragmented data architecture, inconsistent permissions, conflicting definitions, and missing lineage become deal-breakers. Agents don’t “fill in the gaps” like people do. They amplify the gaps. We also explore what an agent-ready architecture looks like in practice: modular interoperability across systems, automated governed access, and shared semantics so every tool and team agrees on what core entities mean. That’s why semantic layers, knowledge graphs, embeddings, and vector databases move from buzzwords to operational necessities. Finally, we talk governance in the agentic era, where systems generate new operational data nonstop, and we lay out a practical way to choose which workflows are worth “agentifying” first. If you’re leading enterprise AI transformation, subscribe for more, share this with a colleague building AI agents, and leave a review with the one workflow you most want to automate responsibly. Want to join a community of AI learners and enthusiasts? AI Ready RVA [https://aireadyrva.com/] is leading the conversation and is rapidly rising as a hub for AI in the Richmond Region. Become a member [https://aireadyrva.com/membership-options/] and support our AI literacy initiatives.

1 de jun de 202612 min
episode Ep 80 - The Competitive Reset: AI Creates New Winners By Moving Value artwork

Ep 80 - The Competitive Reset: AI Creates New Winners By Moving Value

Send us Fan Mail [https://www.buzzsprout.com/2433144/fan_mail/new] AI is everywhere right now: copilots, automated workflows, faster analytics, better dashboards. And yet a lot of leaders still feel the same uneasy question underneath the hype: if AI is so powerful, why aren’t we seeing truly transformational business outcomes everywhere? We dig into the uncomfortable answer: many organizations are solving the wrong problem by treating AI as an efficiency upgrade instead of a shift in competitive dynamics. We unpack the AI productivity paradox and explain why “doing existing work faster” becomes table stakes as tools spread across the market. Using the history of electricity as a clear analogy, we explore why the biggest gains rarely come from swapping in a new tool while keeping the same operating model. The real breakthrough comes when you redesign the system itself: workflows, decision rights, coordination, and how value is created and captured. Then we map three waves of AI value creation: productivity, differentiation, and market restructuring. We talk about AI-native products and experiences, modern AI moats like proprietary data and faster learning loops, and the deepest disruption of all: AI compressing transaction costs and coordination friction across industries. If agentic AI can search, compare, negotiate, and optimize continuously, who wins the customer interface and who gets disintermediated? We close with four strategic questions to help you rethink profit pools, defensibility, learning velocity, and whether you’re redesigning the business or merely automating the old one. If this helped, subscribe, share it with a teammate, and leave a review with your biggest takeaway. Want to join a community of AI learners and enthusiasts? AI Ready RVA [https://aireadyrva.com/] is leading the conversation and is rapidly rising as a hub for AI in the Richmond Region. Become a member [https://aireadyrva.com/membership-options/] and support our AI literacy initiatives.

25 de may de 202615 min