Agentic Conversations (formally mlops.community)

Walking Tokyo Talking Agent Protocols

Two people, a wrong turn into a back alley, a community garden, and about thirty minutes of arguing about protocols on the streets of Tokyo. The guest is Angie Jones, VP of Developer Experience at the Agentic AI Foundation, fresh off launching AGNTCon + MCPCon in China before the Tokyo stop. She opens with what she learned there: a mobile-first, super-app world where the integration problem most of us obsess over barely exists, where every conversation about agents is really a conversation about the model, and where companies are now reaching for MCP and A2A precisely because they want to operate outside that ecosystem.The bulk of it is WebMCP - a protocol with a confusing name and, until recently, almost no attention. The pitch: put tool calling in the page itself, so your agent works inside your logged-in session with only the tools relevant to the page you're on, instead of screenshotting an anonymous browser and burning tokens guessing at the accessibility tree. Angie explains why it went from ignored to urgent the moment agentic browsing got good, and why the fix for computer use being slow and hijacking your machine might be a standard rather than a better model. It closes on agent-to-agent: whether anyone actually wants a marketplace of thousands of agents, or whether the real value is the one agent that has access you'll never get. Plus a well-earned complaint about three-letter acronyms and why researchers are still the only people naming things well. Timestamps:[0:00] Intro[0:59] Launching the conference in China[1:34] What North America gets wrong about agents[2:23] Super apps versus endless integrations[3:14] What happens when they expand beyond China[3:39] Tencent and A2A in production[4:34] A model-first country[5:56] Chinese coding agents and harnesses[6:59] Tokyo and the conference world tour[7:26] What WebMCP actually is[8:56] Why it has nothing to do with MCP[9:20] Page-level tools and your logged-in session[10:36] Why WebMCP sat unnoticed for months[11:25] Token efficiency and reliability[12:23] The moment computer use got good[13:15] Two real grievances with computer use[14:06] Collaborating instead of surrendering your screen[15:24] A short detour into Tokyo signage[16:15] Why web developers should be excited[17:04] Agents and the loss of first-party data[18:27] Why an agent cannot just buy something[19:23] Inside the agentic commerce working group[20:18] Upsells recommenders and an agent that ignores them[23:27] The commerce protocols to watch[24:19] Why A2A is next[25:31] Publishing your agent as a service[27:13] The case against agent marketplaces[27:55] Why access beats capability[29:31] Google's protocol land grab[30:45] Bring back the cool names[31:39] Amsterdam, San Jose, and what comes next

Why Skills And MCP Are Apples And Oranges?

Tool descriptions tell an agent what a tool does. They don't tell it how to use five tools together, in the right order, following your conventions. That gap is where this conversation lives. Filmed at AGNTCon + MCPCon in Tokyo with Ola Hungerford, Principal Engineer for AI Enablement at Nordstrom and a maintainer of the Model Context Protocol, who spent the last several months turning a pattern everyone was quietly reinventing into an actual MCP extension. Ola walks through what skills over MCP really means: the server stops being a pile of tools and becomes a distribution channel, handing the agent the instructions, workflows and knowledge it needs only at the moment it needs them. She explains why server instructions weren't enough, how progressive discovery keeps context from exploding, and why the same mechanism works for memory and preferences even when no tools are involved.Then it gets into the harder parts. What belongs in the MCP spec versus the agent skills spec. Why passing custom front matter through opens a rug pull and prompt injection surface nobody wanted. Where skills start to look like sub-agents, and why there's still no standard way to declare which servers a skill depends on. And the honest problem underneath all of it: how do you standardize something while everyone is still finding out what it's actually for, without breaking a hundred things the next time you change your mind? Timestamps:[0:00] Intro[0:21] AI enablement at Nordstrom[0:33] What skills over MCP actually is[1:29] The MCP server as a distribution channel[2:01] Server instructions versus skills[3:11] Distributing knowledge and memory[4:23] Progressive discovery explained[5:23] Where the idea came from[6:59] From draft to official extension[8:19] What early adopters changed[8:51] Front matter and custom metadata[9:57] Rug pulls and prompt injection risk[10:57] Will any of this get standardized[12:10] Marrying two very different specs[13:03] Skills as personas and sub-agents[13:53] The missing dependency standard[15:15] How the extension actually works[16:19] What harnesses still need to support[16:59] Consent and skill integrity[17:56] Where skills over MCP goes next[18:43] Why cramming 200 tools fails[19:41] Standardizing before you know the answer[22:18] Is git the wrong tool for agents[23:19] Picking tools for the actual persona[23:51] Trying to be less productive[25:48] The anxiety of idle agents[27:15] Why she keeps a robot on her desk[28:22] If the agent feels the friction, does it matter[29:41] Efficiency, waste, and caring enough[30:52] Letting an agent debug for you[32:03] Choosing your rabbit hole

Why Cost Per Million Tokens Is A Useless KPI?

A year ago, Palo Alto Networks built dashboards to track AI spend. Today those dashboards are useless, and the team that built them thinks that's the whole story. Recorded at FinOps X in San Diego, this conversation brings together Abhinav Lad, who leads cloud and AI finance at Palo Alto Networks, and Kuntal Patel, who runs the cloud engineering function behind it. They explain what happened when agents entered the picture, and AI stopped behaving like a service anyone could forecast. The short version: consumption went from linear to exponential almost overnight. Agents are goal-oriented rather than task-oriented, so they plan, call tools, verify, fail, retry, and keep looping until they hit the outcome, and every iteration is billable. So how do you run finance on top of that? Abhinav and Kuntal walk through the metrics that replaced their old forecasts: adoption rate, cost per user, AI as a percentage of revenue - and the budget limits that let engineering leaders choose between the newest model and a longer runway. They get into the open question of whether a cheaper model saves money or just burns more tokens thinking. They explain why an AI gateway became the control plane for cost and security at the same time, why retry caps belong in the design phase instead of the postmortem, and how FinOps starts to resemble product QA once the bill becomes the clearest signal that something is broken. They close on a warning worth sitting with: cost per million tokens is a number that means almost nothing on its own, and a value story built on it will point you somewhere you don't want to go. Palo Alto Networks: https://www.paloaltonetworks.com Abhinav Lad: https://www.linkedin.com/in/abhinav-lad Kuntal Patel: https://www.linkedin.com/in/kuntalpatel35 Alex Salkever: https://www.linkedin.com/in/alexsalkever Timestamps: [0:00] Intro [1:00] Who runs FinOps for AI at Palo Alto Networks [2:10] Last year's AI dashboards are already useless [4:26] Agents turned linear forecasts exponential [7:27] Three traits that make agents expensive [8:34] The hidden bill: RAG, vectors and egress [9:16] Cost per user and adoption rate [11:21] Giving engineering leaders a budget [12:07] Using DORA metrics to prove value [13:53] Where DORA stops fitting AI [16:20] Does the cheaper model actually save money [17:57] Why you need an AI gateway [20:05] Inside Prisma AIRS [21:00] Three cost models for three use cases [22:52] Forecasting lessons from Electronic Arts [24:03] Runaway agents and endless loops [25:59] Capping retries before they burn cash [28:06] Writing cost policy at design time [29:01] When FinOps becomes product QA [32:17] Explaining AI spend to the C-suite [34:51] Valuing AI beyond engineering [37:04] Crawl, walk, run: where they are today [38:20] Why cost per million tokens is meaningless [39:26] Closing thoughts

The Five-Layer Cake Approach to Scaling AI Without Wasting Money

In this episode of Agentic Conversations, we sit down with Ambud Sharma, Principal Engineer at Pinterest, responsible for general technology efficiency, fresh off delivering a controversial keynote on AI infrastructure optimization at scale. Ambud walks us through his Five Layer Cake framework - a structured approach to driving efficiency across every level of the AI stack, from silicon and hardware procurement to model selection, inference engine design, and governance. We explore how decisions compound across layers to unlock real business growth, and how the wrong choices can lock you into expensive commitments for years. We stress test the framework against two very different business models: what the stack looks like if you are building the next Cursor, and how it changes entirely if you are building the next YouTube. Along the way we cover hardware immutability, inference engine warm-up costs, GPU occupancy, context switching, quantization trade-offs, model routing, and why experimentation discipline is the only thing that keeps AI infrastructure costs from getting out of control. We also look at how this framework holds up in the emerging agent era, what changes when agent-to-agent communication becomes the norm, and why agent traffic just passed bot traffic on Cloudflare. The conversation closes on a deceptively simple takeaway: there is no silver bullet, and experimentation at every layer always comes first. Pinterest: https://about.pinterest.com/ Alex Salkever: https://www.linkedin.com/in/alexsalkever Ambud Sharma: https://www.linkedin.com/in/ambud Timestamps: [0:00] Introduction and the controversial keynote [2:09] The five-layer cake explained [4:30] Why hardware decisions are irreversible [6:47] Two business models: building Cursor vs YouTube [10:42] Applying the five layers to a YouTube-style company [14:23] Experimentation as the core efficiency method [17:09] Inference stack: context switching and warm-up costs [20:07] Model layer: why changing models breaks everything [24:10] When you should not use an LLM at all [26:41] Governance and routing: right model for the right task [29:20] Horror stories of unchecked token spend [31:37] Experimentation discipline without stifling innovation [34:35] How the five layers change in the agent era [36:05] Agent-to-agent communication and governance complexity [38:27] Core takeaway: experimentation first at every layer

The Winchester Mystery House Problem in AI Development

AI models are starting to act like appliances, locked into one narrow way of working, instead of the flexible infrastructure they used to be. Drew Breunig, an AI and data strategist working with the Overture Maps Foundation, joins us to explain why, and what it means for anyone building something that doesn't look like Claude Code. Drew walks through his "Winchester Mystery House" idea: what happens once code gets so cheap to write that the only real bottleneck left is feedback. From there we dig into DSPy: signatures, the GEPA optimizer, and the brand-new Flex optimizer, which rewrites your code instead of just your prompt, complete with a real before-and-after on cost and accuracy. We also get into why so many AI-built apps and websites end up looking identical, the actual difference between an agent and a workflow, what Drew learned a year after shipping a code library with no code in it, and why he thinks the most valuable thing you can do right now is close the laptop and go talk to people. CMPND: https://www.cmpnd.ai Drew Breunig: https://www.linkedin.com/in/drewbreunig/ Demetrios: https://www.linkedin.com/in/dpbrinkm Timestamps: [0:00] Cold open: when Claude Code tries to call itself [1:19] Biggest AI news: labs trading diversity for reliability [2:35] How harnesses get trained into models over time [5:41] The problem: your harness starts fighting the model [9:13] When do you need your own harness? [10:02] The Winchester Mystery House warning [16:13] The blank page problem: why everything looks the same [20:50] Infrastructure vs appliances: the thesis lands [22:40] Current tool loadout: GLM, Kimi, Claude Code, Pi [27:04] The Raspberry Pi personal agent running on Slack [31:00] Crystallizing tasks: when to replace AI with pure code [33:10] DSPy explained: separating what from how [35:23] How prompt optimizers actually work [39:31] DSPy pre-dates ChatGPT: model-agnostic programs [44:00] Why you still need to ship the code, not just the spec [50:00] Don't plan more than a month ahead anymore [54:00] Coaching agents all day feels productive — it isn't [57:58] The dopamine of building with agents vs. why you still need human feedback

How Predictive Analytics Stops Budget Overruns Before They Happen?

Every engineer at Wayfair can now see, in real time, exactly what their code costs, and that's on purpose. Brent Eubanks, FinOps Architect at Wayfair, walks us through what happens when you stop treating AI spend as a finance problem and start treating it as an engineering one.The story that sticks with you: a team was burning $400k a month on an LLM-driven workflow, until they flipped the whole thing on its head - hard-coded logic doing the heavy lifting, AI called in only when it's actually needed. New spend: $80k. It's the kind of move that only makes sense once you realize more agentic autonomy isn't always the win it's sold as.From there we get into the machinery Wayfair has built to make cost ownership feel less like a punishment and more like a superpower: guardrails baked straight into coding copilots, predictive alerts that catch a team blowing its budget before the month even ends, a "spend captain" system that pushes budget decisions down to the people closest to the work, and a genuinely strange new question every tech leader is starting to ask: if you were staffing a company with a team of AI agents instead of people, what would that org chart even look like?Wayfair: https://www.wayfair.comBrent Eubanks: https://www.linkedin.com/in/brenteubanksDemetrios: https://www.linkedin.com/in/dpbrinkmTimestamps:[00:00] Intro & teaser [01:14] Meet Brent (FinOps Architect, Wayfair) [02:34] Making AI adoption easy for developers [03:30] Three-area cost governance framework [04:45] Automated FinOps guardrails in pull requests [07:29] Predictive analytics: 400% over budget warnings [08:26] The 400% spender: asset or liability? [10:06] The "spend captain" model [12:13] The goal: 1000x more token use in two years [13:05] The thought experiment [15:08] The $400k workflow rebuilt for $80k [16:42] Hardcoded first, LLM as fallback [19:49] Engineers now own their full cost stack [20:13] The FinOps companion that works overnight [24:10] Guardrails injected before the pull request [26:17] Recurring tasks become automated skills [27:46] Sankey diagram: cost per agent step in real time [28:40] "The bill didn't blow up - that doesn't happen by accident"

How To Delegate To An Agent Like You Would An Employee?

OpenAI's Codex developer experience lead sits down with a former comedian turned ML engineering lead for a conversation about what happens when computer use agents stop being a novelty and start actually running your day. The conversation moves through building an AI-powered morning brief that reads every email, Slack message, and tweet before you've even opened your laptop, letting pinned threads check in on themselves every 30 minutes, and a skills system built to mirror how a person actually writes and reviews code. There's a close look at the guardrails and permission layers that keep an autonomous agent from pushing to the wrong repo or replying to the wrong tweet, how a codebase merging thousands of pull requests a day survives thanks to self-healing review before anything hits CI, and the idea of AI deference - when an agent should push through a task alone versus stop and ask for help. The back half gets personal: why developing taste and vocabulary now matters more than working harder, what it actually takes to delegate to an agent the way you'd onboard a new employee, and why this might be the year voice-orchestrated computer use finally makes everyone feel like they're talking to Jarvis. OpenAI: https://openai.com Monaco: https://www.monaco.com Jason Liu: https://www.linkedin.com/in/jxnlco Mihail Eric: https://www.linkedin.com/in/mihaileric Demetrios: https://www.linkedin.com/in/dpbrinkm Timestamps: [00:00] Intro and guest backgrounds [01:36] Why computer use beats plain API calls [09:11] Building an AI-powered morning brief [10:17] Self-monitoring threads that check in on their own [18:26] How OpenAI reviews thousands of PRs a day [19:42] Self-healing pull requests before CI even runs [23:07] Building review skills from teammates' habits [30:52] Why hard work stops being the differentiator [35:03] Introducing the idea of AI deference [42:44] Learning to delegate like hiring your first assistant [46:22] Why voice beats typing for giving agents context [50:56] The Tony Stark Jarvis analogy for this year

Why Your AI Bill Will Double Before It Gets Better

In this episode, we're joined by Josh Collier, FinOps Lead at Superhuman (formerly Grammarly), to explore what it really costs to run AI at scale and why the rules of the game changed faster than anyone expected. We discuss how AI token costs dropped 80% in two years, why that trend has sharply reversed with frontier models doubling in price, and how Josh rebuilt a single LLM workflow that cost $400k a month down to $80k by rethinking the architecture. He also shares how a cost calculator built in 15 minutes transformed the way his team estimates spend before running experiments, and why research-led optimization is the only kind that works without degrading the product. Along the way, we cover hidden costs most teams miss, the trade-off between Azure reserved capacity and OpenAI Priority Processing, why fixed subscription pricing is broken in an AI-native world, vendor lock-in risk, and what OpenAI's Guaranteed Capacity announcement really signals about where vendor relationships are heading next. Superhuman: https://superhuman.com Josh Collier: https://www.linkedin.com/in/josh-collier-945b7029/ Demetrios: https://www.linkedin.com/in/dpbrinkm Timestamps: [00:00] OpenAI Guaranteed Capacity: what's really going on [01:04] Josh's path into AI FinOps [02:48] Token costs: the 80% price drop [04:16] Why costs will only go up [05:06] External LLMs as financial risk [07:16] Why subscription pricing is dead [08:22] The data residency fee nobody notices [09:33] The cost calculator built in 15 minutes [10:24] How it changed dev team speed [13:00] Tracking costs by service and team [15:33] $400k workflow rebuilt for $80k [17:13] Why only research can optimize tokens [20:00] Speculative decoding win [23:11] One bad query, $40k gone [26:00] Why Azure PTU was exhausting [28:59] Shadow traffic load testing [29:07] Priority processing: no brainer [31:10] Guaranteed capacity: lock-in signal? [32:18] The danger of multi-year AI deals [33:28] Vendor-agnostic proxy as exit strategy

• Transcript

MCP Goes Stateless

David Soria Parra is an Engineering Lead at Anthropic and one of the core maintainers of the Model Context Protocol (MCP). We explore the biggest evolution of the protocol since its launch, and why MCP is becoming the foundation for the next generation of AI agents. We discuss why MCP is moving toward stateless communication, what developers misunderstand about state, sessions, and transport layers, and how lessons from real-world deployments at massive scale have shaped the protocol's future. We also dive into MCP v2, SDK migrations, protocol design, extension architecture, governance, developer experience, and how Anthropic thinks about balancing simplicity with long-term flexibility. Along the way, we explore progressive disclosure, tool search, programmatic tool calling, context bloat, forward compatibility, long-running AI tasks, protocol evolution, open-source governance, observability, and why the future of AI infrastructure will depend on designing protocols that can evolve without breaking the ecosystem. Timestamps: [00:00] Introduction [01:59] Why MCP Had to Become Stateless [04:28] The Tradeoffs of Stateless Design [06:13] What We Learned About Agent State [08:04] Sessions, Models & Implicit State [09:33] Migrating to MCP v2 [12:19] Lessons from HTTP & Open Source Standards [18:16] Shipping Fast Without Breaking Everything [20:35] The Future Complexity of MCP [22:44] Core Features vs Extensions [26:47] Progressive Disclosure Explained [28:16] Solving Context Bloat [30:50] Why Tool Search Beats Progressive Disclosure [32:10] The Biggest MCP Anti-Pattern [34:25] Designing for Forward Compatibility [38:41] Why "Tasks" Matter [40:53] JSON, Tokens & Better Tool Calling [44:44] Observability & Tracing AI Agents [47:34] Will MCP Ever Be Finished? [50:22] What's Next for MCP

• Transcript

AI Hype vs. Real Value

Manish Dasaur is a Managing Director at PwC with over 20 years in data and AI, having helped 100+ clients navigate AI disruption and extract real business value from data, AI, and agentic AI initiatives. In this episode, he breaks down why most enterprise AI programs stall — and the playbook the winners are using instead. Huge thanks to PwC for supporting this episode! 💰 The 30% benchmark — What "good" actually looks like: real efficiency gains clients are reporting across engineering, finance, HR, and supply chain 🔄 Workflows, not use cases — Why isolated pilots and POCs never show up in EBITDA, and how end-to-end workflow redesign does 🧪 Champion vs. challenger — Running a control group against your AI-automated process so ROI is demonstrated, not guessed 📞 Why customer care agents are still freaking hard — Context, CDP integration, billing systems, and voice-to-voice latency 💸 Tokenomics & FinOps — Consumption-based cost surprises, model selection, prompt engineering, and enforcing cost-per-workflow budgets 🔍 Auditing agentic behavior — Using AI to test AI, the missing "SOC 2 for agents," and certifying agents for sensitive use cases 👤 Human in the loop as an evolving scale — From reviewing 50% of outputs down to 10% as trust builds 🧠 88% do AI, 33% scale it — Building a culture of innovation, and why AI usage is showing up in performance reviews 💼 Jobs, reskilling & the operating model reset — Why 75%+ of jobs will be reskilled, not replacedIf you're an AI leader, platform engineer, or exec trying to turn AI experiments into P&L impact, this one's for you. Links & Resources: Connect with Manish: https://www.linkedin.com/in/manishdasaur/ PwC AI: https://www.pwc.com/us/en/tech-effect/ai-analytics.html PwC's 2026 AI Business Predictions: https://www.pwc.com/us/en/tech-effect/ai-analytics/ai-predictions.html Timestamps: [00:00] AI Hype vs Business Value [00:44] API Spend Tracker Widget [02:18] Tokenomics and FinOps for AI [06:19] Measuring AI Impact Objectively [11:07] AI in Support Workflows [18:05] AI Innovation Culture [27:16] MCP Servers and SOC 2 [29:14] Human in the Loop in evolving scale [35:44] AI and Workforce Efficiency [39:59] AI Transformation and Mindset [42:19] Wrap-up

• Transcript

The Creator of FastMCP Explains the Future of MCP

In this episode, we're joined by Jeremiah Lowin, Founder & CEO at Prefect and the creator of FastMCP, to explore how one of the most influential projects in the MCP ecosystem came to be - and where the protocol is heading next. We discuss the accidental origin of FastMCP, why Anthropic adopted it into the official SDK, what developers are getting wrong about MCP, and why Chris believes the biggest opportunity for AI agents isn't customer-facing applications, but internal enterprise systems. We also dive into MCP Apps, developer experience, protocol design, AI tooling, Python, and why building great abstractions is often more valuable than exposing more configuration. Along the way, we explore the rapid growth of the MCP ecosystem, how FastMCP became the default way many developers build MCP servers, why "too much magic" can actually hurt developer experience, and what the next generation of AI-powered applications will look like as agents move beyond simple tool calling into rich, interactive experiences. Prefect: https://www.prefect.io Jeremiah Lowin: https://www.linkedin.com/in/jlowin Demetrios: https://www.linkedin.com/in/dpbrinkm Timestamps:00:00 Lost My Entire Talk00:47 The Story Behind FastMCP02:08 Anthropic Adopted FastMCP02:34 When MCP Took Off04:10 FastMCP vs The Official SDK05:43 Is MCP Actually Dead?06:42 What Everyone Gets Wrong About MCP08:11 MCP's Biggest Use Case10:25 Building Internal AI Systems12:00 Why FastMCP Exploded13:29 Making Complex Software Simple15:10 Can Software Be Too Magical?20:11 MCP Apps Explained23:42 Why Python Needed MCP Apps27:54 The Future of AI Interfaces34:18 AI Should Generate UIs40:11 AI Deleted My Presentation43:30 The AI Assistant We Actually Need48:00 Personal AI vs SaaS52:28 The Future of AI Agents55:06 Final Thoughts

• Transcript

What Happens When Every Developer Has 20 AI Agents?

In this episode, we're joined by Stephen O'Grady, Co-Founder and Principal Analyst at RedMonk, to explore one of the biggest shifts happening in software engineering: AI is making code dramatically cheaper to produce, but everything downstream is becoming the new bottleneck. We discuss why SaaS isn't dead despite the hype, the explosive rise of MCP, why AI agents are overwhelming developer infrastructure, and what happens when every engineer suddenly has dozens of AI developers working alongside them. Stephen explains how package managers, code reviews, security, governance, and enterprise systems are all struggling to keep pace with AI-generated software. Along the way, we dive into AI coding tools, MCP adoption, developer productivity, infrastructure scaling, enterprise software, open source, package repositories, governance, and why the hardest problems in software may no longer be writing code—but managing everything that comes after. RedMonk: https://redmonk.com Stephen O'Grady: https://www.linkedin.com/in/sogrady Demetrios: https://www.linkedin.com/in/dpbrinkm

• Transcript

AI Agents Should Be Treated Like Hackers

In this episode, we're joined by Matt DeBergalis, CTO and Co-Founder of Apollo GraphQL, to explore what happens when AI agents start interacting with enterprise systems that were never designed for them. We dive into the collision between APIs, MCP, GraphQL, and agentic AI, and why traditional assumptions about trust, permissions, and security are breaking down. Matt argues that AI agents should be treated as untrusted actors by default, and explains why giving agents access to enterprise data creates entirely new challenges around governance, access control, and risk management. Along the way, we discuss semantic APIs, enterprise data silos, citizen developers, agent permissions, security boundaries, and how GraphQL and MCP can work together to make enterprise systems more accessible to both humans and AI. The conversation also explores why companies are racing to deploy agents despite the risks, and what the future of enterprise software might look like when AI becomes the primary consumer of APIs. Apollo GraphQL: https://www.apollographql.com Matt DeBergalis: https://www.linkedin.com/in/debergalis Alex Salkever: https://www.linkedin.com/in/alexsalkever Timestamps: [00:00] AI, APIs, and Trust [01:16] MCP API Lessons [06:16] GraphQL and MCP Integration [12:55] API Security for MCP [16:10] Linux Kernel Security Concerns [19:09] API Design and Controls [21:52] Trust in Autonomous Systems [25:06] MCP GraphQL Wish List [27:13] API Access Patterns [28:44] GraphQL API Perspective

• Transcript

Developers May Stop Depending on Libraries

In this episode of Agentic Conversations, we're joined by Shaun Smith, software engineer, open source advocate, and contributor at Hugging Face, to explore how AI coding has changed almost overnight. We dive into reinforcement learning, MCP (Model Context Protocol), Fast Agent, Claude Code, open source AI, and why today's language models have become so capable that many traditional software libraries are becoming "liquefied." Shaun explains how reinforcement learning unlocked long-running autonomous agents, why ideas are becoming more valuable than code, and how developers should think about building software in an era where AI can generate entire applications. Along the way, we discuss Hugging Face's MCP server, Fast Agent, AI-powered developer tools, multimodal applications, MCP Apps, context windows, coding assistants, Rust, Python, TypeScript, open-weight models, software architecture, and what the future of programming looks like when humans increasingly focus on design instead of implementation. Shaun Smith: https://www.linkedin.com/in/smithshaun Demetrios: https://www.linkedin.com/in/dpbrinkm Hugging Face: https://huggingface.co ⏱️ Timestamps[00:00] Introduction [01:56] The State of Open Source AI [05:18] Reinforcement Learning Changed Everything [07:50] Fast Agent Explained [10:18] Fast Agent as an MCP Reference Platform [12:20] Building Smarter AI Tools at Hugging Face [15:17] Natural Language Search Instead of APIs [17:46] Why MCP Apps Matter [20:06] The Evolution of MCP Apps [23:05] Building AI-Native User Interfaces [26:12] Context Is the New Programming Language [28:00] The End of Code Libraries [29:50] Why Developers Aren't Writing Code [31:25] AI Changes Software Engineering [33:05] The Future of Open Source AI [35:43] Claude Skills That Save Hours [38:02] Training Models with AI [39:05] Building Your Own AI Tools [40:50] MCP for Consumers, Enterprises, and Developers [43:42] Why Shell Access Makes Agents Smarter [45:18] Secure Agent Workflows [46:08] The Future of AI Interfaces [47:02] Outro #HuggingFace #MCP #OpenSourceAI

• Transcript

10 Cities. 4 Countries. One Unexpected MCP Lesson.

In this episode, we're joined by Ben Morss, Developer Advocate at DeepL, who spent months traveling across North America and Europe teaching developers about MCP, building MCP servers, and helping teams understand how AI agents actually use tools. We dive into the biggest misconceptions around MCP, why so many developers still misunderstand how it works, and what Ben learned after giving talks and workshops in 10 cities across four countries. Along the way, we explore MCP server design, tool calling, security concerns, translation workflows, developer education, and how DeepL is using MCP to bring high-quality language translation into AI-powered applications. DeepL: https://www.deepl.com Ben Morss: https://www.linkedin.com/in/ben-morss-ph-d-15bab15/ Alex Salkever: https://www.linkedin.com/in/alexsalkever Timestamps: [00:00] AI and API Integration [00:41] DeepL at DevSummit [01:19] MCP Roadshow Origins [03:47] MCP Hackathon Insights [07:52] Security in Model Protocols [10:25] AI Expert vs Noob Queries [16:08] DeepL vs Frontier LLMs [18:16] MCP vs REST API [21:39] MCP Servers and DeepL

• Transcript

The Next Programming Language Is English

In this episode, we're joined by Cornelia Davis, Developer Advocate at Temporal and a longtime software architect who has spent decades helping shape modern cloud-native systems. We explore how programming has evolved from assembly language to cloud-native architectures, and why AI is forcing us to rethink software development once again. Cornelia argues that natural language is becoming a new programming abstraction, while durable execution may be the missing layer that makes AI agents reliable in production. The conversation dives into probabilistic software, long-running AI agents, MCP tasks, human-in-the-loop workflows, durable timers, distributed systems, and why developers may no longer need to think about infrastructure the way they once did. Cornelia Davis: https://www.linkedin.com/in/corneliadavis Demetrios: https://www.linkedin.com/in/dpbrinkm Temporal: https://temporal.io Timestamps [00:00] AI Programming Abstractions [00:52] Abstraction Evolution in Programming [04:05] Text to SQL Evolution [10:08] Compensations for Natural Language [12:13] Durable MCP in AI [18:34] Streaming Session Explanation [21:31] Batch Processes with Tasks [29:29] Complexity Relocation in Systems [33:10] Complexity Relocation in Dev [36:36] Programming Model Shifts

• Transcript

Omnigent: Composition, Control, and Collaboration for AI Agents

Denny Lee is PM Director, Startups & Ecosystem at Databricks, a longtime Apache Spark, MLflow, and Delta Lake contributor — and one of the people behind Omnigent, the open-source meta-harness Databricks just released under Apache 2.0. He joins Demetrios to explain why the industry is moving from models to harnesses to meta-harnesses, why token spend is replaying the CapEx-to-OpEx shift all over again, and why he's using debating AI agents to plan a matcha farm in Taiwan. In this episode: 🍵 Agents as research partners — Denny uses dueling agents to scout matcha-growing regions in Taiwan, down to soil pH, elevation, and processing infrastructure 🥊 Why agents should debate each other — letting two models argue surfaces the questions you didn't know to ask 🔱 Forking conversations — the missing UX pattern: branch a session, keep the shared context, explore two threads in parallel 🧠 The meta-harness layer — how Omnigent sits above Claude Code, Codex, Pi, and custom agents so models and harnesses become hot-swappable parts 👥 The two-pizza rule for agents — military span-of-control logic says you can manage 5–7 agents before you lose the thread 💸 Tokenomics is the new DevOps — the CapEx→OpEx playbook repeats: give developers spend visibility, keep central governance for the rest 🛡️ Policies, budgets, and guardrails — enforcing cost caps and approval rules at the harness layer instead of inside prompts 🤖 Auto model selection — why classic machine learning (not another LLM) may be the right way to route tasks to cheap vs. frontier models ✍️ "Created by" vs. "assisted by" — the open source accountability debate: whoever submits the code owns the code 🗄️ Databases are back — agents need cheap, stateful memory, which is why Postgres, Lakebase, and serverless databases are having a moment If you're building with coding agents, managing AI spend, or trying to keep up with the harness arms race, this one's for you. Links & Resources: Omnigent (open source): https://www.databricks.com/blog/introducing-omnigent-meta-harness-combine-control-and-share-your-agents Omnigent GitHub: https://github.com/databricks/omnigent Denny Lee on LinkedIn: https://www.linkedin.com/in/dennyglee Denny's blog: https://dennyglee.com Tokenomics Foundation announcement: https://www.finops.org/insights/finops-x-2026-day-1-keynote/ Timestamps: [00:00] SOA to LLMOps Transition [01:06] Agentic Research Workflow [10:45] Agent Debate for Execution [13:53] Agentic Footnote Concept [24:41] Harnesses in Agent Systems [32:43] Harnessing Multi-Layered Agents [38:06] Token Spending Awareness [41:01] Token Spend Efficiency [43:53] Model Selection Frustration [51:06] Meta Harness in AI [53:15] Harness Layers Model [57:17] Wrap up #Tokenomics #AIAgents #Omnigent

• Transcript

The Current State of Agentic Retrieval - Qdrant Roundtable

Qdrant Roundtable episode: The Current State of Agentic Retrieval Join the Community: https://go.mlops.community/YTJoinIn Get the newsletter: https://go.mlops.community/YTNewsletter MLOps GPU Guide: https://go.mlops.community/gpuguide Big shout-out to Qdrant for the collaboration! // Abstract AI agents are only as good as the information they can find, retrieve, and remember. In this community roundtable with the Qdrant team, we explored the latest advances in agentic memory, vector search, retrieval systems, and production AI architectures. As AI agents move beyond simple chatbots into systems that can reason across large amounts of information, retrieval is becoming one of the most important layers in the AI stack. The discussion covered the real-world challenges of building agents that remember what matters, forget what doesn't, and consistently retrieve the right context at the right time. If you're building AI agents, RAG systems, or production AI applications, this conversation offers practical insights into where retrieval is headed and what it takes to build reliable, scalable agentic systems. // Bio Ewa Szyszka Ewa is a Developer Relations professional based in San Francisco with a background in Computer Science and Hardware Engineering, passionate about bridging the gap between technology and the developer community. She holds a BSc in Computer Science and an MSc in Electronics, bringing a strong blend of deep technical foundations and communication skills to her work. Dylan Couzon Dylan is based in New York City, and he helps developers build better AI applications. He is passionate about AI, programming, open source, and robotics, and enjoys sharing what he’s building and learning along the way. Neil Kanungo Neil is an experienced professional with expertise in data science, developer relations, and product growth. Currently serving as the Head of Developer Relations at Qdrant, Neil previously held the position of VP of Product Led Growth & Developer Relations at KX, where significant increases in product registration and user activation were achieved. At TIBCO, Neil managed a team focused on enhancing the adoption of TIBCO Spotfire through various initiatives, including tutorial videos and live webinars. With a strong technical background, Neil has developed innovative solutions in analytics, machine learning, and data visualization across multiple roles, including Engineering Data Analyst and Asset Integrity Engineer at Enterprise Products. Neil holds a Bachelor of Science in Radiation Physics from The University of Texas at Austin, a Master of Science in Mechanical Engineering from Texas Tech University, and is pursuing a Master in Applied Data Science from the University of Michigan. Evgeniya Sukhodolskaya Developer Relations at Qdrant with 8 years of IT experience across software engineering, machine learning, and technical management, and 4 years in Developer Relations. Holds a Master’s in Machine Learning, Data Analytics, and Data Engineering. Passionate about NLP, data-centric AI, and the role of vector search in advancing AI technologies. Andrei Cristea Andrei is a Berlin-based Developer Relations Engineer at Qdrant, a prominent open-source vector database. With a Master’s degree in Artificial Intelligence from TU Munich, his expertise bridges AI, data infrastructure, and knowledge engineering. Hosted by Demetrios // Related Links Website: https://qdrant.tech/ ~~~~~~~~ ✌️Connect With Us ✌️ ~~~~~~~ Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore Join our Slack community [https://go.mlops.community/slack] Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)] Sign up for the next meetup: [https://go.mlops.community/register] MLOps Swag/Merch: [https://shop.mlops.community/]

• Transcript

AI Agents in Healthcare?

Kingsley Madikaegbu is the founder of HealID, a startup building agentic AI on top of the Model Context Protocol (MCP) for one of the most heavily regulated environments there is: healthcare. Recorded at MCP Dev Summit North America in New York, Kingsley sits down with Alex Salkever of the Agentic AI Foundation to break down how you give patients, doctors, caregivers, and family members each their own agent over the same medical record — without breaching HIPAA, leaking PHI, or letting an agent quietly go off the rails. In this conversation:🏗️ The four-layer architecture — Dumb data at the bottom, then access permissions, then MCP, then reasoning agents on top. Why logic never touches the data layer.🔐 MCP vs REST — Why enforcing per-role compliance in a REST API meant encoding permissions everywhere, and how MCP collapses that mess.🪪 HIPAA, auditability & traceability — Proving a specific person (not a snooping agent) accessed a record, with a full audit trail that regulators actually accept.🎟️ The nightclub-bouncer analogy — How MCP reorganizes the entire "club" per guest instead of just checking a VIP list.⌚ Wearables & real-world data — Turning an Apple Watch arrhythmia signal into a triaged, severity-scored workflow with doctors in the loop.🧭 Deterministic vs model-driven — Why anything clinical or regulatory stays binary, and the agent-as-coach (not decision-maker) pattern for patients.🛑 Keeping agents on the leash — Tool restriction, behavioral metadata, and drift/anomaly detection so an agent can't reinterpret its own job.⚡ The instant kill switch — Revoke permission, and the agent returns a hard 404, never partial data.⚖️ The liability question — When an agent follows a designed workflow and something goes wrong, who's responsible: patient, host, or provider? The industry hasn't decided.📋 Kingsley's MCP wishlist — Built-in traceability (OTEL-style spans), native time-bound enforcement, and guardrails against agent-to-agent data leakage.If you're building agentic systems for healthcare, finance, legal, or any regulated industry where "the agent did it" isn't a good enough answer — this one's for you.Links & Resources🔗 HealID — https://gethealid.com/🔗 Kingsley Madikaegbu — https://www.linkedin.com/in/kmadikaegbu🔗 Alex Salkever / Agentic AI Foundation — linkedin.com/in/alexsalkever🔗 MCP Dev Summit North America — https://events.linuxfoundation.org/mcp-dev-summit-north-america/Timestamps:[00:00] Intro[00:13] AI Agent Liability[01:10] MCP in Healthcare AI[06:30] MCP vs REST Architecture[11:29] Healthcare Integration Challenges[18:29] Non-compliant Patient Challenges[24:13] Deterministic vs Model-Driven Workflows[28:08] AI in Healthcare Conversations[34:38] Agent-to-agent workflows in healthcare[38:02] Future MCP security

• Transcript

Coding Agents Are Secretly General Agents

In this episode: 🧠 Coding agents are generalist agents — why "positive transfer" means an agent that's better at code is better at everything, and how that makes them "AGI-complete" ⏳ "Code will be solved in a year" — what the automation of knowledge work actually looks like, and why Jay joined ClickUp to be on it 🏗️ Why the labs are crushing AI startups — free-for-two-years deals, Windsurf losing Claude access, and the brutal economics of building on top of frontier models 🔗 The real moat is convergence — context, surfaces, and unit economics, a.k.a. "Cursor for your whole job" 💬 Slack's data walls & the Glean problem — why fragmentation is the enemy and a single system of record wins 🧪 RLVR & verifiability — why code became the perfect training ground for agents, and how to tell if you're even getting better 🔬 LLMs are running the frontier of science — Putnam 12/12, Erdős problems, simulating a cell, and vibe-writing economics papers 🚗 The car wash test that still breaks GPT-5 — spiky models, world models, Plato's cave, and the "stochastic parrot" debate 🏖️ Plus: mechanistic interpretability as "brain surgery," catastrophic forgetting, the danger of deleting knowledge from models, and a pitch for a "resort for LLMs" Whether you're building agents, leading an AI team, or just trying to figure out what "agentic" really means for everyday work — this one's a fun, deep ride. 🔗 Links & Resources Jay Hack: linkedin.com/in/jayhack ClickUp: clickup.com MLOps Community: go.mlops.community Mentioned: Gödel, Escher, Bach (Douglas Hofstadter) · "Machine Learning: The High-Interest Credit Card of Technical Debt" (Sculley et al.) · Periodic Labs · Ginkgo Bioworks · Physical Intelligence ⏱️ Timestamps [00:00] AI Timeline [00:22] AI Startups and Timing [06:30] GPT-3 Impact [13:24] Selling CodeGen to ClickUp [19:31] AI Interaction Patterns [28:41] Slack and AI Agents [36:11] ClickUp Task Automation [41:32] AI in Scientific Research [48:48] Human Understanding vs AI [54:18] Catastrophic Forgetting Explained [59:59] AI Delegating to Humans [1:05:00] Agent-Based Game Integration [1:08:42] LLM vs Game Design [1:11:27] Wrap up #AIAgents #AgenticAI #ClickUp

• Transcript