Munich AI Conference

Amazon Bedrock gives organisations access to a broad choice of foundation models through a single managed service. This session covers how Bedrock is used in Europe: model availability in EU regions, data residency and sovereignty, and the security and compliance controls that matter for regulated industries. Expect practical guidance on taking generative AI workloads from experiment to production on AWS.

On 30 September 2025 a national broadcaster’s site died at three million concurrent viewers — and the analytics died with it. Why reactive autoscaling loses that race, and what we built in eleven working days instead, without moving a single server.

A hands-on deep dive into two of AWS's newer AI offerings. Amazon Quick brings AI agents for research, business intelligence and automation to knowledge workers, while Kiro is an agentic IDE for spec-driven software development. Sascha shows both tools in action and where they fit in an organisation's AI toolset.

Qbit42 is a German no-code platform for building AI assistants and agents, with data stored in Germany and GDPR-compliant processing in Europe. Thomas demonstrates how business teams can connect documents, data and systems to create agents without writing code. A key focus is model independence: keeping tools and workflows intact when the underlying model changes.

As agents take on longer, multi-step tasks, memory and context management become key drivers of both quality and cost. This session shows how Semvec, an AI memory infrastructure layer, can be integrated into an agent harness with MongoDB as the operational data platform. The talk walks through the architecture and how structured memory improves agent output while reducing the tokens sent to the LLM.

EggAI develops an open-source, async-first meta framework for building enterprise-grade multi-agent systems. Damian explains how agents built with different frameworks can communicate and collaborate through a shared messaging layer. The deep dive covers architecture patterns and what it takes to run multi-agent systems reliably in production.

MAIKERS is one of Munich's largest AI communities, connecting practitioners, founders, researchers and corporates. Johannes shares what the community sees on the ground: which AI topics are moving into real use, what teams are building and where collaboration across the ecosystem creates value. The session closes with an outlook on upcoming MAIKERS formats and how to get involved.

OpenAI develops the GPT model family and ChatGPT, used by a large and growing number of people and enterprises. Tilman gives an overview of current developments in OpenAI's models and platform and how organisations apply them in day-to-day work. The talk highlights patterns from enterprise deployments and what they mean for teams planning their next AI initiatives.

Successful automation depends on more than technology. This panel discusses why people need to be designed into AI-driven processes from the start, from change management and upskilling to human oversight and accountability. Bringing perspectives from consulting, IT services and industrial manufacturing, including experience in responsible AI, the panellists share what it takes to bring employees along on the AI journey.

RP Intelligence advises small and medium-sized enterprises on AI and digital transformation, from strategy through implementation to training. Drawing on a decade of experience at large consultancies, Stefan shows how mid-sized companies can identify relevant AI use cases and implement them with tools such as the Power Platform and RPA. The focus is on human-centred adoption and measurable business results.

TrendAI is Trend Micro's enterprise business, renamed in 2026 to focus on security for AI-driven organisations. As companies deploy models, agents and new data flows, their attack surface changes. Marco outlines the key risks emerging around enterprise AI and how security teams can protect AI applications while also using AI to strengthen their own defences.

Devoteam is a European tech consultancy focused on cloud, data, AI and cybersecurity. Stefan shows how operational intelligence, combining real-time operational data with AI, helps organisations move from isolated use cases to AI embedded in daily operations. The talk covers the building blocks from data foundations to decision automation, along with lessons from transformation projects.

New Relic's observability platform gives engineering teams full-stack visibility into applications and infrastructure. Klaus presents New Relic's observability frameworks and how agentic AI helps teams detect, analyse and resolve issues faster. He also looks at the other side: how to monitor AI applications and agents themselves in production.