QCon Software Development Conferences

APIs for Agents: Rethinking API Programs in the MCP Era Model Context Protocol (MCP) is shifting the API landscape. We’re moving from static documentation for developers to dynamic discovery for autonomous agents. At scale, traditional "manual" governance becomes a bottleneck. At QCon London 2026 (March 16-19), James Gough, Distinguished Engineer, API Platform Lead Architect @Morgan Stanley, and Andreea Niculcea, Vice President @Morgan Stanley, eplore how to evolve API programs for the agentic era without drowning in bureaucracy. The implementation reality: The agentic shift: Rethink the "API product" when the primary consumer is an autonomous agent rather than a human developer. Governance at speed: Learn to introduce critical governance changes and roll out fixes across 200+ APIs without slowing delivery to a crawl or creating endless upgrade work. Scalable guardrails: Build self-service guardrails that empower teams to move fast while keeping security outcomes measurable and reducing manual coordination. Connect with API leads navigating similar governance evolutions. The shift to agentic consumers is happening now, and the patterns are still emerging. Early bird pricing ends February 10: https://bit.ly/4k8R8FP #QConLondon #APIDesign #MCP #AIEngineering #SoftwareArchitecture

20 years of vetting technical signals from noise. Most conference talks are polished success stories or product launches in disguise. For two decades, QCon has focused on the opposite: No sponsors on the main tracks. No hidden sales pitches. Just the unfiltered trade-offs of systems that survived production traffic. Registration is now open for QCon San Francisco (Nov 16–20, 2026). Our 20th-anniversary program features 12 provisional tracks, including: - AI Engineering that Delivers - Navigating Major Architecture Migrations - The Stories Behind the Incidents - Architecting for Efficiency Beyond the sessions, the real value comes from the whiteboards and unconference sessions where you can share experiences and pressure-test designs with senior peers facing similar scale constraints. Secure launch pricing until February 10th. Groups of 10+ can access additional savings on the full conference experience. 🔗 https://bit.ly/3LO4xqs

Stop shipping AI prototypes. Start building AI systems. Most AI demos work in a vacuum. Scaling them to handle production traffic, security constraints, and cost optimization is where the real engineering begins. At QCon AI Boston (June 1-2), we skip the high-level hype to focus on the technical requirements of enterprise-ready AI. The 2026 implementation roadmap: Architect for scale: Moving beyond basic RAG to build resilient, distributed AI platforms. Operational reliability: Deep dives into MLOps, LLM observability, and GPU cost optimization. Secure governance: Implementing guardrails that ensure compliance and security without stalling velocity. SDLC acceleration: Practical patterns for using AI to increase engineering throughput across the entire lifecycle. Led by Practitioners: Learn from a Program Committee that operates at scale, including Eder Ignatowicz @RedHat, Meryem Arik @Doubleword, and Hien Luu, @Zoox, Author of MLOps with Ray. Scaling AI is a cross-functional challenge. Bring your architects and leads to align your roadmap and reduce implementation risk, and learn from peers who've already navigated these decisions. Register here: https://bit.ly/46oAg8q #QConAI #AIEngineering #MLOps #SoftwareArchitecture #ProductionAI

Ever wonder how Netflix Commerce systems evolved over the last 10+ years of the company growth in both scale and needs? I was there for much of the journey and will share some lessons at #QConLondon in March. It's my first time speaking at a European conference and I'm very excited! https://lnkd.in/ghS9PFuP

The right 300 tokens beat 100k noisy ones: The architecture of context engineering When an AI agent "forgets" an instruction from three messages ago, the problem isn't the window size; it’s the architecture of your data retrieval. At QCon London 2026 (March 16-19), Patrick Debois, AI Product Engineer @Tessl and Co-Author of the "DevOps Handbook", moves the conversation from prompt engineering to context engineering. The technical deep-dive: Beyond RAG: Dissecting when to use embeddings, when to leverage Model Context Protocol for precision, and when a simple lookup is the superior choice. Solving context drift: Architectural patterns to ensure your model stays focused on the signal, even as the conversation scales. Evaluation over vibes: Building data-driven eval strategies to prove your context choices actually improve model reliability. Your context window is a limited resource. Learn how to manage it as an architectural component rather than a text box. Learn alongside practitioners wrestling with the same retrieval trade-offs. Early bird ends February 10: https://bit.ly/4q4mq2h #QConLondon #AIEngineering #LLMOps #MachineLearning #GenerativeAI

Your tests verify the code works. But how do you verify the design works? Unit tests and type systems can only verify that the pieces of your system work. They can't tell you if the architecture itself is fundamentally broken. By the time a design-level bug hits production, the cost to fix it is 10x higher than a simple refactor. At QCon London 2026, Hillel Wayne, Author of Logic for Programmers, explains how to use Formal Methods to find resilience bugs in systems that don’t even exist yet. Why this matters for your next sprint: - Identify flaws early: Learn to "debug" your architecture on paper before you write a single line of code. - Property-based thinking: Move beyond the "happy path" to define the properties your system must have (and the ones it definitely shouldn't). - Distributed systems reality: A practical framework for verifying concurrency and consistency in complex, distributed environments. Discuss practical application with architects facing similar distributed systems challenges at QCon London 2026. Early bird pricing ends February 10. Explore the session: https://bit.ly/4kscuhX #QConLondon #SoftwareArchitecture #DistributedSystems #FormalMethods #SystemDesign

Moving AI from cool demo to production-grade. While the hype cycle continues, senior engineers are busy solving the actual infrastructure challenges of shipping AI at scale. At QCon London 2026 (March 16-19), we’ve skipped the fluff to focus on how AI is actually being operationalized in complex environments. Here are just two examples of the sessions focused on the implementation reality: Beyond Context Windows: Karthik Ramgopal, Distinguished Engineer @LinkedIn, breaks down how to move past brittle heuristics to build persistent memory and state management for agentic systems. Rewriting All of Spotify’s Codebase, All the Time: Jo Kelly-Fenton and Aleksandar Mitic, Software Engineers @Spotify, explain how Spotify is using LLMs to handle the "boring" but critical work of managing large-scale software migrations and codebase maintenance. Instead of high-level theory, these are deep dives into the trade-offs of agentic memory, cost-efficient scaling, and automated refactoring. Join architects and leads navigating the same production AI challenges. The hallway conversations alone are worth the trip. Explore AI-focused sessions at QCon London: https://bit.ly/46gR4y3 Early Bird pricing is available through February 10. #QConLondon #LLMOps #AIInfrastructure #SoftwareArchitecture

Looking forward to speaking at QCon. Governance that protects you without slowing you down is a particular challenge with the rapid adoption of AI. Many more tools to assess and to integrate into your workflow!

I’m thrilled to be hosting a track at the 20th anniversary of #QConLondon this March. My goal was to bring together practitioners who are actually solving challenges in taking AI applications to production. If you’re a senior engineer navigating these challenges, the Architectures in the Age of AI track is for you. Join us March 16-19! You can use SPEAKER60QUK26 for £60 off. https://qconlondon.com/

Skip the research phase. Benchmark your roadmap in 3 days. Validating a technical strategy through fragmented blog posts and case studies takes months. At QCon London 2026 (March 16-19), you get the alternative: direct insights from practitioners who have already shipped the architectures you're currently evaluating. Why senior engineers use QCon as a shortcut: 🔹 Validated patterns: Stop guessing. Compare your stack against production-grade implementations of AI, platform engineering, and modern data architectures. 🔹 The "unwritten" docs: Get the implementation details that aren't in the README 🔹 Accelerated execution: Use "lessons learned" from your peers to move from research to production without the standard trial-and-error. 🔹 Technical judgment: High-signal sessions designed to help you distinguish between stable patterns and high-maintenance hype. Don't wait for the documentation to catch up. Avoid architectural dead-ends by learning from those who have already navigated them. Early bird ends February 10: https://bit.ly/4t0yQL0 #QConLondon #SoftwareArchitecture #PlatformEngineering #SystemDesign