Upgrade to Pro — share decks privately, control downloads, hide ads and more …

ABCS25: Platform Engineering for the Digital Ag...

ABCS25: Platform Engineering for the Digital Age by Romano Roth

⭐️ Transforming DevOps with Azure: Platform Engineering for the Digital Age#
In today’s rapidly evolving business environment, organizations face unprecedented challenges: rapidly shifting customer demands, intensifying competition, and a growing shortage of skilled professionals. Agility and speed have become critical to survival. This talk will showcase how Platform Engineering, powered by Azure, serves as the foundation for overcoming these challenges. By automating complex workflows and infrastructure provisioning, it delivers the scalability and flexibility needed to adapt swiftly, innovate faster, and address skill gaps effectively. Through in-depth insights, real-world examples, and a live platform demo, this session will explore how Azure-driven platform strategies can transform development processes, enhance productivity, and enable the rapid delivery of high-quality software. Attendees will discover actionable techniques for overcoming scaling challenges, improving developer experience, and streamlining operations by leveraging DevOps practices within a robust, Azure-integrated platform.
🙂 ROMANO ROTH ⚡️ Global Chief of DevOps @ Zühlke

Tweet

More Decks by Azure Zurich User Group

Other Decks in Technology

Transcript

  1. Romano Roth Chief of Cybernetic Transformation & Partner Transforming DevOps

    with Azure Platform Engineering for the Digital Age
  2. 2025 Roth, Romano © Zühlke My passion: helping companies bringing

    people, processes, technology and AI together, so that they can deliver ongoing value to their customers. Romano Roth Chief of Cybernetic Transformation & Partner [email protected] https://www.romanoroth.com/ @RomanoRoth Romanoroth RomanoRoth DevOps Meetup Zürich DevOpsDays Zürich CAS Digital Transformation CAS Enterprise Architecture CAS DevOps Leadership and Agile Methods HSLU HSLU HSLU
  3. 2025 © Zühlke Roth, Romano Me trying to keep up

    with the AI news and products launches
  4. 2025 © Zühlke Roth, Romano Vibecoding was introduced in February

    2025, and there are already books about it… 05.06.25 | 7 Vibe coding (or vibecoding) is a programming paradigm dependent on artificial intelligence (AI), where a person describes a problem in a few sentences as a prompt to a large language model (LLM) tuned for coding. The term was introduced by Andrej Karpathy in February 2025. Vibe coding. Wikipedia. Retrieved May 6, 2025, from https://en.wikipedia.org/wiki/Vibe_coding © Zühlke 2025
  5. 2025 © Zühlke Roth, Romano Who has let an AI

    develop a complex system on its own? (like an air traffic control system)?
  6. 2025 © Zühlke Roth, Romano Who believes that AI will

    one day develop and operate complex systems on its own?
  7. 2025 © Zühlke Roth, Romano AI investments vs. revenue AI

    investments are exploding. But where’s the revenue? Who’s going to pay for the gap? Investment: 2023: $ 200B 2024: $ 600B 2025: $ 1’000B* Revenue: 2023: $ 80B 2024: $ 100B 2025: $ 120B* https://www.sequoiacap.com/article/ais-600b-question/ https://www.sequoiacap.com/article/follow-the-gpus-perspective/ * Estimation based on internet research
  8. 2025 © Zühlke Roth, Romano My Opinion: We’re not just

    surrounded by hype. We’re ruled by it. The Dot-com bubble, Friday March 10, 2000 We need an online shop! The AI bubble, ? 2025 We need an AI thing!
  9. 2025 © Zühlke Roth, Romano We are living in the

    era of the AI Idiot! Where ChatGPT cowboys, clueless politicians, and short-term hype managers burn millions on buzzwords. Disclaimer: I have purposely overstated this statement. Of course, I have the utmost respect for real experts like this group of people and especially for the engineers, architects, and serious researchers who engage deeply with this topic." or transport it on the talk track.
  10. 2025 © Zühlke Roth, Romano …and yet I think AI

    is awesome I use it every day It is a fantastic TOOL
  11. 2025 © Zühlke Roth, Romano We need to cut through

    the fog and focus on what truly matters.
  12. 2025 © Zühlke Roth, Romano AI won’t fix your broken

    Processes, Organization, Technology, or Governance
  13. 2025 © Zühlke Roth, Romano Cybernetic means building adaptive systems

    where processes, organization, technology, and governance operate in feedback loops to continuously improve and support the company’s business goals.
  14. 2025 © Zühlke Roth, Romano The key to a Cybernetic

    Enterprise is having the right foundation.
  15. © Zühlke 2025 Public Roth, Romano Industrialization of Software Engineering

    26 Platform Local Dev Local Dev Production Local Dev Digital Product Digital Product Digital Product Product Team Product Team Product Team Production Digital Product Digital Product Digital Product Product Team Product Team Product Team Digital Product Digital Product Digital Product Digital Product Digital Product Digital Product Platform internal and with partners Distributed and heterogeneous Development Environments internal and with partners Products ¡ Kubernetes (AKS) ¡ Crossplane ¡ Kafka ¡ Ingress NGINX ¡ MQ: RabbitMQ ¡ API Gateway: Tyk, Gloo ¡ DB: MS SQL, MySQL, Oracle, PostgreSQL, MongoDB, MariaDB, Redis, ElasticSearch, Neo4j ¡ S3: MinIO ¡ IDP: Azure Entra, Keycloack ¡ GitLab ¡ JFrog ¡ ArgoCD ¡ Hashicorp Vault ¡ Prometheus ¡ Grafana ¡ Open Telemetry ¡ Trivy Services ¡ Kyverno ¡ Tempo ¡ Calico ¡ Aqua ¡ SonarQube ¡ Jenkins ¡ Cypress ¡ Playwright Huge Tool Landscape Standards of preselected Services and Products and Templates (paved pathes) internal from to from to Source: CNCF mixed external internal mixed external
  16. © Zühlke 2025 Public Roth, Romano Product Owner Scrum Master

    BA UX/UI FE Dev BE Dev QA DevOps-Eng Security-Eng Automation-Eng Infra-/Cloud-Eng Product Owner Scrum Master BA UX/UI FE Dev BE Dev QA Product Owner Scrum Master BA UX/UI FE Dev BE Dev QA Self Service Platform Tech Stack Tech Stack Developer Experience Platform Team Product Team Product Team Product Team Platform Engineering Target Operating Model Target Operating Model for Product Development in the Cybernetic Enterprise 27 Tech Stack Product Owner Scrum Master BA UX/UI FE Dev BE Dev QA DevOps-Eng Security-Eng Automation-Eng Infra-/Cloud-Eng Product Owner Scrum Master BA UX/UI FE Dev BE Dev QA DevOps-Eng Security-Eng Automation-Eng Infra-/Cloud-Eng Product Owner Scrum Master BA UX/UI FE Dev BE Dev QA DevOps-Eng Security-Eng Automation-Eng Infra-/Cloud-Eng Product Team Product Team Product Team Multiple Autonomous Product Teams § Complexity § Cognitive team overload § Heterogeneity Platform engineering is the discipline of building and operating self-service internal developer platforms.
  17. © Zühlke 2025 Public Roth, Romano Your Company Platform Team

    Product Team A Product Team B Product Team C Product Team N Your Cybernetic Platform Application Runtime/ Compute Access & Identity Developer Experience Automatic DevSecOps Centralized Security Observability GitOps … … Platform Engineering Enables DevOps in Product Teams | 28 Platform Team develops, builds and maintains the platform Product Teams Practice DevOps to build, run and maintain products, using the platform Platform Team Delivers the Platform as a Product to Enable DevOps Generates Value for Customers Generate Value for your Teams 05/06/2025 Adapted from: Richard Watson, Use Platform Engineering to Scale your DevOps Practices Across People Process Technology and Culture, Gartner, 2022
  18. 2025 © Zühlke Roth, Romano OK, but how does that

    look like in practice? 05/06/2025 | 29
  19. © Zühlke 2025 Public Roth, Romano 05/06/2025 | 30 …

    GitLab Artifacts Vault Repository GitHub HashiCorp Automation Blocks Artifacts Repository Directory Azure AD Automation Blocks CD ArgoCD TBD Your Tool Runs on Creating awesome products Developer Compliance & Security CISO Tracking Cost Program Manager On/off boarding Head of Development Your Company Standardization Enterprise Architect • Faster Time to market • More value for money • Better quality • Better Customer satisfaction • Better Employee satisfaction … UX BA / RE PO SM Architect • Application Runtime/Compute • Developer Experience • Automatic DevSecOps • Access & Identity • GitOps • Observability • Centralized Security K8s K8s Block TBD TBD Packages K8s Catalog Hypervisor Cost OIDC Admin QA Platform Integrated Tools Internal Developer Portal Provisioning & Automation (Cross Blocks / Tools) Unified Integration Blocks Adapters & Connectors Unified API Example: High Level Architecture of the Cybernetic Platform at Zühlke
  20. © Zühlke 2025 Public Roth, Romano Hub Spoke Network Architecture

    | 31 DDoS Protection Azure Firewall + partner Policies Virtual WAN Hub Network Space #1 Spoke Network Space #1 Kubernetes Peering Space #2 Spoke Network Space # 2 Kubernetes Peering Space # 2 VMs Developer Notebook § The Hub is the central point of connectivity beetween all spokes and cross-premises § A central firewall controls all traffic (ingress/egress). Central policies apply (managed by partner) § The firewall monitors & inspects traffic and logs to Azure Sentinel (SIEM, SOAR) § The Hub and its resources are in a separate Azure Subscription and managed by partner fully automatized Space #3 Spoke Network Space # 3 Kubernetes Peering
  21. © Zühlke 2025 Public Roth, Romano The Key: Floating platform

    A Platform integrates best-in-class tool chains for software product development in a highly efficient way 32 Characteristics of a Platform ¡ Standing platform team and community able to navigate a rapidly changing tool environment and select the best tools ¡ Highly efficient, standards-based integration of commercial and open-source tools into the platform ¡ Covering the entire range of functions across all relevant subject areas ¡ Excellent developer experience through easy use via portal and self-service catalog ¡ Supports tool life-cycle management via a managed service catalog ¡ Utilizes the innovative power and speed of the entire tool and platform provider ecosystem, without lock- in risks Tools DevOps Platforms Floating platform Market view Red Hat GitHub Azure Others PORTAL / SELF SERVICE CATALOGUE DEVELOPER EXPERIENCE Visualisation is illustrative and does not show real coverage Adapted from Gregor Hohpe (2024). Platform Strategy Cloud
  22. © Zühlke 2025 Public Roth, Romano How does that work

    in a bank? Delivering High Quality Software Requires the Right Platforms and Tools 33 Workload Placement Developer Placement Cloud OnPrem Location Ext. Location Developer-Cloud Cloud-Environments (TE/IN/PR) Software Development Lifecycle / Environments Development Test Integration Production DevSecOps DV TE IN PR Shift Left Platform Platform Ext. Developer Int. Developer Int. Developer Platform P6 P5 P4 P3 P2 P1 P6 P5 P4 P3 P2 P1 Continuous Deployment Pipeline § Continuous Integration § API-Mocks § Synthetic Test data De-Coupling Layer (API’s) Satellite System Satellite System OnPrem Environments (TE/IN/PR) Core Banking System Azure, AWS, GCP
  23. © Zühlke 2025 Public Roth, Romano Your Company Platform Team

    Product Team A Product Team B Product Team C Product Team N Your Cybernetic Platform Application Runtime/ Compute Access & Identity Developer Experience Automatic DevSecOps Centralized Security AI / ML … … … AI is a capability of your Cybernetic Platform | 35 Platform Team develops, builds and maintains the platform Product Teams Practice DevOps to build, run and maintain products, using the platform Platform Team Delivers the Platform as a Product to Enable DevOps Generates Value for Customers Generate Value for your Teams 05/06/2025 Adapted from: Richard Watson, Use Platform Engineering to Scale your DevOps Practices Across People Process Technology and Culture, Gartner, 2022
  24. © Zühlke 2025 Public Roth, Romano Your Cybernetic Platform Offers

    AI Capabilities “as a Service” to the whole company Platform Interfaces GenAI infrastructure components – compute, network, storage Apps Tools Models Chatbots AI coding assistents Generative design tools Synthetic data tools Knowledge management Productvity assistants Prompt Engineering RAG and fine-tuning tools GenAI application frameworks Vector databases Model life cycle management (DataOps, MLOps, ModelOps) AI TRiSM (AI trust risk and security management) tools Model hubs (app store for models) Visibility into model usage and cost Sandbox environments for experimentation Open-Source Models Enterprise-specific models Domain-specific models Resuable GenAI components (eg. Summarization, content synthesis and analysis, prediction) Common generative AI Platform capabilities Developer Portal Conversational CLI API Product Development Teams Adapted from How Platform Engineering Teams Can Augment DevOps With AI, Manjunath Bhat, 16.01.2024, Gartner
  25. © Zühlke 2025 Public Roth, Romano Where are we moving?

    The future belongs to those who master the symphony of organization, processes, technology, and AI FROM TO
  26. © Zühlke 2025 Public Roth, Romano Outcome-Oriented Product Leadership §

    Align teams to long-term value over short- term delivery § Prioritize impact through evidence-based decision-making § Combine product discovery and delivery into one continuous cycle Architecture Customer-Centric by Design § Embed user research across discovery and delivery § Prioritize inclusive, outcome-driven design § Validate assumptions early through experimentation Flow-Optimized Delivery System § Organize around value streams for clear ownership § Enable continuous planning with agile cadences § Deliver small, testable increments to reduce risk Insight-Enabled Decision Making § Data is the nervous system of the enterprise § Enable real-time visibility through observability platforms § Drive actions through telemetry, leading indicators, and adaptive metrics Empowering Through Platforms § Self-service platforms accelerate team autonomy § Shift-left quality through automation and observability § CI/CD pipelines enable continuous innovation Architect for Adaptability and Speed § Modular, API-first systems for composability § Built-in scalability, observability, and resilience § Evolvable architectures aligned to value flow Engineering Data Agile Programme Delivery Product management Customer experience Cybernetic Enterprise A Cybernetic Enterprise is a continuously evolving, self-regulating system that learns through feedback, adapts with intelligence, and aligns purposefully to deliver lasting value in an ever-changing world § Insight-Driven: Real-time observability, telemetry, and feedback loops power evidence-based learning and decision-making § Continuously Adaptive: Systems and practices evolve in response to dynamic environments through intelligent feedback and experimentation § Self-Regulating by Design: Autonomous teams operate with high alignment through principles, not control, enabling decentralized execution § Systemically Aligned: People, technology, and processes form a coherent, purpose-driven ecosystem across domains § Flow-Oriented: The entire enterprise is structured around the end-to-end flow of value, minimizing delays and maximizing outcomes Core Characteristics
  27. © Zühlke 2025 Public Roth, Romano Summary AI-Augmented DevOps with

    Platform Engineering Platform Engineering We are in a hype 00 01 02 Build your floating Cybernetic Platform Build your Cybernetic Enterprise AI Capabilities as a service 03 04
  28. © Zühlke 2025 Public Roth, Romano We are entering the

    age of Cybernetic Enterprise The future belongs to those who master the symphony of organization, processes, technology, governance and AI. This is the way how you build your Cybernetic Enterprise! | 41