The AI Track at AWS Summit Hamburg 2026: From Demo to Deployment

AWS Summit Hamburg 2026 AI Track

Last year, I wrote about the dedicated Gen AI track at AWS Summit Hamburg 2025. The response was overwhelming — the track was packed, conversations spilled into the hallways, and the Fischbrötchen at the Landungsbrücken afterwards sealed the deal. Hamburg won me over.

This year, the AI track is back — bigger, sharper, and with a clear theme: from demo to deployment. If 2025 was about showing what generative AI can do, 2026 is about making it work in production. And the track reflects that shift.

What’s Different This Year

Last year’s track covered the breadth of generative AI — from RAG basics to creative content generation. This year, the focus has narrowed to what matters most right now: agentic AI in production. Almost every session addresses a piece of the puzzle that separates a working demo from a system you’d trust with real customers.

The track features 15 sessions across all levels (200–400), 6 customer stories from companies like Deutsche Bahn, Delivery Hero, Siemens, Infineon, HUK-Coburg, and Würth, and a mix of breakout sessions, chalk talks, and hands-on labs.

The Sessions — Organized by What You’re Trying to Solve

Getting Started: The Big Picture

Getting started with the AI track

AIM201: From demo to deployment: solving agentic AI’s toughest challenges (200, 09:00) This is my session — I wrote about it here. I’ll walk through the top challenges builders face when moving AI agents from prototype to production — build vs. buy decisions, reliability, observability, cost, security, and evaluation. If you attend one session to frame your day, this is the one. I’ll also point you to the deeper sessions for each topic.

AIM202: Bridging from POC to production: An intro to Amazon Bedrock AgentCore (200, 12:30) The natural sequel to my talk. Siemens shares their experience using AgentCore’s Runtime, Gateway, Identity, Memory, and Observability services to move from proof-of-concept to production.

Production Challenges: The Hard Parts

Production challenges in agentic AI

AIM301: AI Agents at Deutsche Bahn: Amazon Bedrock AgentCore best practices (300, 13:30) This is the governance session. Deutsche Bahn is running agent-to-agent collaboration with shared memory, identity-based access controls, and MCP gateways — at scale. If you’re wondering how to manage dozens of agents across teams without chaos, this is your session.

AIM306: Improve agent quality in production with Bedrock AgentCore Evaluations (300, 14:30) How do you know your agent is getting better, not worse? This session covers pre-built metrics, custom criteria, CI/CD integration, and live trace sampling for production evaluation.

AIM311: Architecting scalable agentic AI with Bedrock AgentCore at Infineon (400, 14:30) The 400-level deep dive. Secure runtime, identity, policy enforcement — with Infineon’s customer story. For the architects in the room.

Trust and Security

Trust and security for AI agents

AIM304: Building Trust and Explainability into AI with Automated Reasoning (300, 12:30) Formal verification to reduce hallucinations. AgentCore Policy for fine-grained tool access control. Live demos.

AIM309: Red Teaming LLMs — Practical Defense Strategies (300, 11:30, AWS Labs) A chalk talk on prompt injection, jailbreaking, and secure code execution. Hands-on and interactive.

Cost and Data

Cost optimization and data for AI

AIM302: Menu Intelligence at Scale: Delivery Hero’s $600K+ Savings Story (200, 11:30) 90% reduction in LLM spend. 85% faster time-to-first-token. $600K+ in annual savings. Prompt caching, LLM-as-Judge, and automated quality control. The cost optimization story of the track.

AIM310: Transforming AI storage economics with Amazon S3 Vectors (300, 14:30) S3 Vectors: native vector storage with up to 90% cost reduction. HUK-Coburg’s production deployment story. If you’re running RAG at scale, this one’s for you.

AIM305: Build Enterprise AI Apps Faster: Bedrock’s Multimodal Solutions (300, 15:30) Documents, audio, images, video — unified in one service. Air’s success story.

The Cutting Edge

Cutting edge AI sessions

AIM313: Using Strands Agents to build autonomous, self-improving AI agents (300, 09:00) Agents that identify knowledge gaps, self-modify reasoning, and build their own tools. If you want to see what “build custom” looks like with the latest open-source framework.

AIM402: Agent Experience: REST in Peace — Why LLMs Can’t CRUD (400, 12:30, AWS Labs) A provocative chalk talk on why REST APIs fail for agent orchestration and what intent-based APIs look like. For API designers and platform engineers.

AIM308: Build more effective agents through model customization (300, 16:30) Nova customization on SageMaker AI for agentic use cases. When the base model isn’t enough.

AIM303: Build a well-architected foundation for scaling generative AI (300, 15:30) N26’s story of building the platform that supports all AI applications across the organization.

AIM307: Würth Lens: AI-Powered Visual Search at Scale (200, 16:30) 10,000+ business users, scaling internationally. A different kind of AI story — visual search for industrial products.

My Top 5 Picks (If You Can Only Attend Five)

  1. AIM201 (09:00) — Start here. I’ll frame the challenges and point you to the deep dives.
  2. AIM301 (13:30) — Deutsche Bahn’s governance story is the most advanced agent deployment in the track.
  3. AIM302 (11:30) — Delivery Hero’s $600K savings story is the proof point every CFO needs.
  4. AIM306 (14:30) — Evaluation is the unsexy work that separates production from demo. This session makes it practical.
  5. AIM402 (12:30) — The most thought-provoking session in the track. REST APIs weren’t designed for agents.

About the Event

Whether you’re building your first agent or scaling to hundreds, the AI track has something for you. See you in Hamburg.

Stefan Christoph is a Principal Solutions Architect at AWS, focused on AI/ML and Media & Entertainment. He writes about AI agents, developer productivity, and the path from demo to deployment at schristoph.online.