The Good Thing

How to Make Your APIs Work for AI | Erik Wilde (Jentic)

Learn more here [https://wundergraph.com/the-good-thing/32-erik-wilde] Talk to a WunderGraph expert here [https://wundergraph.com/contact/sales] Erik Wilde is a legend in the API space. He's had a monumental impact on how many of us think about API design, governance, and digital evolution today. Erik spent many years at Axway where he worked closely with some of the largest organizations in the world on API strategy and governance. Today, he is an OAI Ambassador at the OpenAPI Initiative, runs his own YouTube channel “Getting APIs to Work”, and continues to help organizations align APIs, business capabilities, and real-world outcomes. At Jentic, Erik is now focused on a question many teams are struggling with: "How do you make large, complex API landscapes actually work in an AI-driven world, without breaking everything you already have?" In this episode, Erik joins us for an honest conversation about APIs, AI readiness, and why most enterprises are still thinking about API design at the wrong level. We dig into what breaks when agents meet real-world API landscapes, why fine-grained APIs do more harm than good for AI, and how intent-driven API orchestration is the missing layer between legacy systems and the future of intelligent automation. Hosted by Stefan Avram and Jens Neuse, this episode blends deep technical insight with real-world enterprise experience, covering everything from API design and governance to AI bubbles, agent architectures, and why most API problems are organizational problems, not technical ones. In this conversation, we get into: - Why API design and governance still matter more than ever, especially in the age of AI - What “AI-ready APIs” actually mean, and what most teams get wrong - Why many AI projects fail before they even get started - Why fine-grained APIs overwhelm agents and LLMs - Why narrow REST APIs are the wrong interface for AI, and what should sit on top instead - What it really takes for startups to earn trust with large enterprises If you have ever worked with APIs at scale, struggled with governance, or wondered how AI really fits into existing enterprise systems, you’ll enjoy this one. - HOW TO FIND ERIK Find Erik Wilde and "Getting APIs To Work" here: https://www.youtube.com/@ErikWilde Erik's LinkedIn: https://www.linkedin.com/in/erikwilde/ Visit The Good Thing on YouTube (WunderGraph - Stefan & Jens) [https://www.youtube.com/@WunderGraph] to watch our episodes! Watch this episode here: How to Make Your APIs Work for AI | Erik Wilde (Jentic) [https://youtu.be/lfbNRHOOOIM] Visit us at wundergraph.com [https://wundergraph.com/]

The Man Behind HubSpot's AI Integrations & MCP | Cody Holden

What makes an API truly AI-ready? Cody Holden is the Senior Product Manager for MCP and the Integrations Platform at HubSpot, where he’s been leading a major overhaul of HubSpot’s API architecture and building their MCP client for agentic use from the ground up. Before HubSpot, Cody worked across journalism, technical writing, and product leadership, giving him a rare, multidisciplinary perspective on communication, developer experience, and platform design. Cody joins Stefan Avram for a deep dive into the next generation of APIs, integrations, and AI protocols, and how companies can prepare for an agentic future. Here’s what you’ll learn: From Writing to Product Leadership Cody shares how starting his career in journalism and technical writing shaped his approach to product management, communication, and developer empathy. Inside HubSpot’s API Evolution We break down how HubSpot modernized its APIs, what it means to build “AI-ready” systems, and the governance challenges that come with operating at HubSpot’s scale. MCP: Hype vs. Reality Some claim MCP will replace APIs. Others say it’s just another tool. Cody explains where MCP actually fits in the integration landscape and what unique problems it solves. Becoming a Better Product Manager Cody explains what separates great technical PMs from the rest, how the role is changing in the AI era, and the key habits he’s developed across his career. Looking Ahead From agentic workflows to new API paradigms, we discuss where the ecosystem is headed over the next decade and what platforms will need to embrace to stay relevant. If you’re building APIs, managing integrations, or navigating the intersection of AI and developer platforms, this episode is for you. Jump into the comments: - What does “AI-ready” mean in your organization? - Will MCP become part of every integration strategy? Visit The Good Thing on YouTube (WunderGraph - Stefan & Jens) [https://www.youtube.com/@WunderGraph] to watch our episodes! Watch this episode here: The Man Behind HubSpot's AI Integrations & MCP | Cody Holden [https://www.youtube.com/watch?v=AhxEELOI5XY] Visit us at wundergraph.com [https://wundergraph.com/]

Responding to the Rumors

No, we’re NOT abandoning Federation. We’re taking it to the next level. This week on The Good Thing podcast, Stefan Avram (CCO) and Jens Neuse (CEO) break down how Federation reshapes collaboration, scales developer velocity, and sets the stage for the next era of intelligent systems. From GraphQL to REST, gRPC, and beyond, they explain why a “Federation Layer” is now essential infrastructure for modern engineering teams. We’ll start with the basics: what Federation actually is, what problems it solves, and why it’s central to WunderGraph’s vision, before diving deep into how we’re evolving the model with the Entity Layer and Cosmo Connect. Here’s what you’ll learn: Federation 101: The Big Picture We set the record straight: Federation isn’t dead, it’s just misunderstood. We’ll deep-dive into why Federation remains core to WunderGraph’s roadmap. Expect a brutally honest look at how the shift from monoliths to supergraphs changes how teams build, deploy, and ship software, and why Federation should be about connecting people, not just services. The Entity Layer: The Real Magic Jens explains why this layer is the biggest leap forward since the birth of GraphQL Federation. We unpack how the Entity Layer gives LLMs real business context, why it’s the missing link between human collaboration and machine intelligence, and how it redefines what “federated” even means. Cosmo Connect and the Future of Federation Forget GraphQL-only Federation: the future speaks gRPC. Stefan and Jens dive into why they’re betting big on multi-protocol Federation, and why clinging to GraphQL as the only solution limits innovation. We’ll explore how Cosmo Connect unifies REST, SOAP, and gRPC into one Federated model, and why that’s essential for modern, AI-ready architectures. AI Meets Federation: From Toy Projects to Real Systems Everyone’s talking about LLMs, but few know how to make them production-ready. We’ll break down how Federation gives AI systems context, consistency, and governance, stopping them from becoming just another demo project. Rebuilding Federation from Scratch If we could burn down Federation and rebuild it tomorrow, what would stay, and what would go? The hosts share candid lessons from years in the trenches, what the community gets wrong, and how Federation 2.0 (and beyond) will look when built with people, not protocols, in mind. If you’re building APIs, running microservices, or exploring AI-ready architecture, this isn’t a tutorial. It’s a manifesto for what’s next. Join us in the live chat or in the comments! We’d love to hear: How is your team approaching Federation today? Is GraphQL still central, or are you going multi-protocol? If you’re averse to adopting Federation, why? -- Chapter Markers 00:00:00 What this episode is about 00:02:52 Why WunderGraph is all-in on Federation 00:05:02 The Working Group 00:07:47 eBay partnership and real-world Federation problems 00:09:58 Subgraph authorization 00:11:27 Semantic Non-Null and upcoming products 00:14:24 MCP, AI, and Federation as the backbone 00:21:22 What the Entity Layer is and why it matters 00:32:14 Why next-gen Federation speaks gRPC (Cosmo Connect) 00:43:14 Summit reactions: APM templates, operators, connectors 00:49:06 We’re hiring; retreats, meetups, and community plans 00:52:55 AI hype vs API fundamentals: founder perspective 00:57:22 Vertical AI + API design/governance strategy 01:03:11 Closing + what’s next New episode every other week here and on YouTube (WunderGraph - Stefan & Jens) [https://www.youtube.com/@WunderGraph]! Watch this episode here: Responding to the Rumors [https://youtu.be/z4gWO2QhAaM?si=WkwtYugSwudRAM6Z] Visit us at wundergraph.com [https://wundergraph.com/]

Scaling Support Without Burning Out Engineers (ft. Viola Marku)

Is great customer support as important as a great product? Viola Marku is the Customer Success Manager at WunderGraph, where she partners closely with engineering-led teams to drive successful adoption of federated architecture across complex organisations. With nearly a decade of experience in technical startups, she brings both empathy and precision to every customer engagement, translating abstract challenges into clear, actionable paths forward. Viola's cross-functional work with engineering and sales has helped reduce time to value while elevating the role of customer success as a strategic growth function. She's focused on creating the conditions for customer ownership, embedding success into how modern platform teams actually work. Hosted by WunderGraph CEO Jens Neuse and COO Stefan Avram, they'll talk support excellence, feedback loops, and share advice they’ve learned along the way. Here’s what you’ll learn: Support is a Product * Great support isn't an afterthought. Viola and Stefan make the case that support is part of the product, and why investing early pays off. From Chaos to EOC and beyond * We trace the evolution of WunderGraph’s support process, from Stefan juggling sales and support alone, to the launch of the Engineer-On-Call (EOC) system in 2025, to how Viola reshaped workflows when she joined the team earlier this year. Turning Feedback into Features * How WunderGraph collects, prioritizes, and acts on customer feedback, and why this feedback loop is one of their biggest product advantages. Failures, Fixes, and Lessons * We talk about real mistakes: the support tickets that went sideways, the growing pains, and how even the worst moments became learning opportunities. Scaling a Support Culture * With a growing user base, how will WunderGraph maintain its customer-first values? From AI tooling to culture building, we explore how the team is preparing for what’s next. Whether you’re scaling a devtool, leading a support team, or just curious how real customer obsession shapes product growth, this episode is for you. Jump into the comments or live chat! We want to hear: * Is great customer support as important than the product itself? * What’s the biggest support fail you’ve learned from? * Can great support be your moat? Tune in every Friday at 9 AM on YouTube (WunderGraph - Stefan & Jens) [https://www.youtube.com/@WunderGraph] for new Live episodes! Watch this episode here: Scaling Support Without Burning Out Engineers — Viola Marku [https://youtu.be/6DtxzFY8hnI] Visit us at wundergraph.com [https://wundergraph.com/]

API Leader Daniel Kocot — How to Build Great APIs

APIs run the modern digital world, but what separates the good from the great? Daniel has been part of the codecentric team since October 2016. Since the beginning of 2022 he works as Principal API Consultant at the Dortmund branch. Starting as a consultant with a focus on application lifecycle management, his focus shifted more and more towards APIs. In addition to numerous customer projects and his involvement in the open source world around APIs. He is also a frequent speaker at conferences. In this episode of The Good Thing with Stefan & Jens, Daniel joins us for a deep dive into how modern teams should think about APIs: as products, as capabilities, and as core business enablers. From governance models to open source adoption to the future of API standards, this conversation explores what it really takes to make APIs work at scale. Here’s what you’ll learn: From APIs to Capabilities Daniel explains why thinking in terms of capabilities (“Ship Order”, “Process Payment”, “Approve Loan”) instead of technical endpoints can reshape API design. We discuss how this mindset shift ensures APIs align with business value and how product thinking drives long-term success. Governance vs. Gateways APIs don’t succeed without governance. Daniel shares why tools like API gateways are helpers, not leaders. Together, we explore strategies for balancing autonomy with control, defining ownership, and preventing API sprawl without killing innovation. Open Source as the API Backbone codecentric deliberately builds on open-source technologies. Daniel, Jens, and Stefan discuss how communities like OpenAPI, AsyncAPI, and GraphQL accelerate progress, the trade-offs of open source in enterprise settings, and why standards are crucial for event-driven APIs. The Future of APIs What role will GraphQL, gRPC, and AsyncAPI play in shaping tomorrow’s API strategies? How might AI influence API design and usage? Daniel shares his perspective on when to use different approaches, where they converge, and why reliability still matters most. If you care about API strategy, developer experience, or building infrastructure that lasts, this conversation is for you. Join the live chat or comments and share: * Should APIs be designed as capabilities, not endpoints? * How much governance is too much? * What’s the role of open source in API strategy today? Tune in every Friday at 9 AM on YouTube (WunderGraph - Stefan & Jens) [https://www.youtube.com/@WunderGraph] for new Live episodes! Watch this episode here: How Great Teams Build Great APIs — Daniel Kocot (API thought leader) [https://youtu.be/f5CJYcxIo1A] Visit us at wundergraph.com [https://wundergraph.com/]