The future belongs to companies that not only test AI, but integrate it productively, securely and economically into their processes. The way to get there leads through modern AI-native platforms, clear governance, digital sovereignty and a new generation of intelligent agent systems. In this issue, we show how companies are successfully scaling Agentic AI, keeping their costs under control and creating the technical foundation for the next stage of their digital transformation. Look forward to new solutions, technical insights, community events and practical experiences from the world of cloud native, platform engineering and AI engineering.
Have fun reading, your white duck team wishes you!
AI Operations & Governance
Token Control 5.1: Governance for everything agents do
With Token Control 5.1, organizations gain greater visibility and control over AI applications and agents. New governance capabilities for Responses API, Web Search, File Search, and MCP support compliance, security, and cost control across all Azure AI models. Test now easily and conveniently in the SaaS model, also including AI model deployment.
Microsoft Foundry Agents under control with token control
Microsoft Foundry enables the rapid deployment of AI agents. Token Control adds a single layer of governance for budgets, limits, compliance, and cost control across multiple environments.
Token Control for ISVs: Selling AI Features Instead of Subsidizing
AI functions are easily integrated, and billing is often the bigger challenge. Token Control enables consumption-based allocation of costs and budgets per customer, user or feature and creates the basis for monetization and FinOps.
AI Engineering: Discipline saves more than any model change
It is not the model that determines the costs of an AI solution, but the handling of context, agents, and development processes. We show which measures companies can implement immediately to sustainably reduce their AI costs.
Agentic AI
Flock v0.5.500: Enterprise Security for Agentic AI
With Flock v0.5.500, Agentic AI is ready for enterprise use. New security features such as Entra-ID authentication and guardrails against prompt injection increase protection and governance for productive agent systems.
Flock & Dapr: Agent Swarms That Survive Reboots
Productive agent systems require more than powerful models. The integration of Flock and Dapr provides key building blocks for resilient, scalable, and cloud-native agent platforms.
The live demo in the Dapr Community Call on Youtube.
AI pilots are not enough: Why SMEs need an AI Foundation
Many companies have successfully implemented the first AI pilots. However, for productive use, an AI Foundation is needed that takes governance, security and scalability into account from the outset.
Company News
Wrap-up: Cloud Native Summit Munich 2026
More than 400 participants, 82 speakers and 67 sessions once again made the 5th Cloud Native Summit Munich the meeting place for the cloud-native and AI community. Topics such as Agentic AI, platform APIs and digital sovereignty were the focus.
Save the date – CNS Munich 2027 findet am 14. & June 15, 2027 statt.
You can buy the tickets now and/or become our event sponsor. All details are available on the website .
Our team continues to grow
Our team is getting further reinforcement: With Dylan Ciletti and Eva Kallrath, we are expanding our expertise in the areas of software engineering and platform engineering.
Dylan brings experience in C#/.NET, Python and in the modernization and optimization of existing software systems.
Together with Eva, we are also strengthening our platform engineering team and supporting our customers in building modern, scalable cloud and AI platforms.
More about Eva & more about Dylan.
Events
Webinar ReLive: After the hype comes the craft
AI-supported development promises a lot of potential. In the lecture, Nico will show what experience companies have already gained and which measures are really relevant now.
Wrap-up: ContainerDays – Hijacking agents, without any exploit
How can modern AI agents be compromised without exploiting classic software vulnerabilities? The key takeaways from Nico’s live demo can be found in the presentation slides and the GitHub repository.
If you want to experience the talk live, Nico will give the talk at the Cloud Native Days Austria on 29.09.26. All details available at the following link.
To the slides & to the GiHub Repo.
AI Meetup Wasserburg: Plug & Play for AI Agents
Agent Skills package skills as reusable building blocks for any agent. Martin Brandl, CTO of white duck, will show in short demos on September 24 at 7 p.m. how to find, install and use skills directly, even without a developer background.
IAMCP TechTalk: The Context Layer for Enterprise Agents: Work IQ MCP Servers
How do AI agents gain access to relevant company knowledge? In TechTalk, we show how the Work IQ MCP Server makes information from Microsoft 365 available to your own agents. More on this on October 22 in TechTalk.
After the Hype: How ISVs Are Turning Agentic AI into a Product
The AI hype is over, now it’s the implementation that counts. Together with ADN, we will show how software manufacturers are using Agentic AI productively and accelerating their development processes with GitHub Copilot. Join us in Munich on November 11 .
AI White Sausage Breakfast: When the Agent Changes Sides
Security is becoming the decisive factor for Agentic AI. At the next AI Weißwurst Breakfast, we will discuss current attack vectors, agent hijacking and zero trust strategies for AI agents. Join us on November 13 in Kufstein.
Good to Know
Ingress NGINX is discontinued – this is how the migration to AKS succeeds
The end of Ingress NGINX presents many Kubernetes teams with important architectural decisions. We compare the most important alternatives for AKS and show suitable migration paths for different requirements.
