Two days, built for engineers.
Sessions, keynotes, and breaks across every stage. Released day by day as the lineup locks in.
Where to be, and when.
- 09:30 – 10:15CupertinoSpec-Driven Development: Making AI Coding Assistants Build What You Actually WantSoftware Architecture & Engineering ExcellenceRoland Huß
- 10:30 – 11:15RedmondPipeline Patterns and Antipatterns - Things your Pipeline Should (Not) DoCloud, DevOps & Platform EngineeringDaniel Raniz Raneland
- 10:30 – 11:15Los GatosKnowledge is the infrastructure. Everything else is just tooling.AI Engineering & DataDaniel Ostrovsky
- 10:30 – 11:15CupertinoWhy Your Best Engineer Makes Your Worst First-Time Manager?Engineering careersAleksandra Lemańska
- 10:30 – 11:15Palo AltoDesigning Reliable Distributed Systems: Failures, Retries & IdempotencyCloud, DevOps & Platform EngineeringVioletta Pidvolotska
- 11:30 – 12:15CupertinoDealing with eventual consistencySoftware Architecture & Engineering ExcellenceDennis van der Stelt
- 11:30 – 12:15Los GatosDebugging Intelligence: How Do You Debug a System That Is Thinking?Cloud, DevOps & Platform EngineeringNishant Gupta
- 11:30 – 12:15RedmondFrom Infrastructure Metrics to Business Impact: Measuring Reliability Where It MattersCloud, DevOps & Platform EngineeringTomasz Szarek
- 11:30 – 12:15Palo AltoBeyond the Codebase: Giving Coding Agents the Business Context They're MissingAI Engineering & DataRaj Navakoti
- 12:30 – 13:15CupertinoUnderstanding at machine speed with thousands of contextual toolsAI Engineering & DataTudor Girba
- 12:30 – 13:15Los GatosSearch as Code: The Infrastructure Behind Long-Running AgentsAI Engineering & DataAleksandr Nikolenko
- 12:30 – 13:15Palo Alto"AI: The Whole Company, or Nothing"Engineering careersAgur Jõgi
- 13:30 – 14:15CupertinoAgents as a Workload: Platform Engineering for Autonomous SystemsCloud, DevOps & Platform EngineeringAiswarya Venkitesh
- 13:30 – 14:15Los GatosMagical Mystery Tour: A Roundup of Observability DatastoresCloud, DevOps & Platform EngineeringJosh Lee
- 13:30 – 14:15Palo AltoThe Missing Paper Trail for Agentic EngineeringAI Engineering & DataRizèl Scarlett
- 13:30 – 14:15RedmondBuilding an AI Inference PlatformAI Engineering & DataMichał Dojlidko
- 14:30 – 15:15CupertinoThe Typing Got Cheap. The Judgment Didn't: An Infrastructure Engineer's Year with AI AgentsAI Engineering & DataSachin Malhotra
- 14:30 – 15:15Los GatosForward-Adopt: Shipping Product Today on Tomorrow's PlatformSoftware Architecture & Engineering ExcellenceMisha Kazakov
- 14:30 – 15:15RedmondDesigning the ultimate software engineering checklistSoftware Architecture & Engineering ExcellenceAntonio Olmo Titos
- 14:30 – 15:15Palo AltoHow Hellofresh Group orchestrates agent teams for marketing outcomesAI Engineering & DataAlan-Michael Khalil
- 15:30 – 16:15CupertinoIntentional ReliabilityCloud, DevOps & Platform EngineeringNiall Murphy
How Hellofresh Group orchestrates agent teams for marketing outcomes
Marketing teams are increasingly surrounded by AI tools, but connecting those tools into a reliable operating system for real business outcomes is a much harder problem. In this talk, we'll explore how HelloFresh Group built an agentic marketing platform that coordinates specialized agents across the ad lifecycle: understanding creative briefs, analyzing performance and audiences, recommending concepts, generating new ads, preparing assets for activation, and applying human approval before actions. Rather than relying on one general purpose chatbot, the system uses a supervisor agent that routes work to focused capabilities. Each capability has its own instructions, tool permissions, and operating boundaries. This allows the platform to combine creative intelligence, performance data, asset libraries, Ad generation, and ad-operations systems while keeping spend-related decisions governed and reviewable. We'll look at the engineering patterns behind that orchestration: - capability-based tool permissions instead of unrestricted agent access; - durable conversation state and resumable human-in-the-loop workflows; - structured handoffs between agents, rather than passing ambiguous prose; - background watchers for long-running external jobs; - Slack-based approval and notification flows; - observability across model calls, tools, traces, and outcomes; - safeguards against duplicate generation, accidental launches, runaway polling, and stale state; - and a user experience designed around "coworking" with an agent team The goal is to create a dependable system that helps marketing teams move from insight to creative action, while preserving human judgment, operational control, and a clear record of what happened. You'll leave with a practical view of how to design multi-agent systems for marketing: where to use autonomy, where to add constraints, and how to turn a collection of AI capabilities into a governed team that can deliver measurable outcomes.
Save your seat before it fills.
Early pricing runs while the programme is still being finalised.