Agile Meets Architecture Conference, Day 1 Recap

Yesterday marked the opening day of the Agile Meets Architecture Conference in Berlin, and it was a fitting start for a gathering that aspires to bridge the gap between hands-on agility and high-level architectural thinking. As one of the Program Chairs, I had the privilege of welcoming experts from across the Domain-Driven Design, Software Architecture, and Team Topologies communities. Although the conference is relatively small, its intimate scale encouraged a lot of exchange, candid conversations, and a spirit of genuine collaboration. This atmosphere set the stage for a day filled with thought-provoking talks, culminating in a high-energy Team Topologies Panel that left me and many others with new perspectives on how to optimize the interplay between agile delivery and robust architectural practice.

Below, I’ve summarized the talks I attended. Before we get there, I want to highlight my personal key insight of the day. It came from Simon Wardley’s keynote: “The future of system architecture is all about deciding who makes what decisions and where.” I am currently writing and researching about staffing, procurement, make / buy / customize strategies für fast flow environments where I pull in various perspectives such as DDD domain classifications, Cynefin and Wardley Maps. Simon brought up the perspective of decision making in the context of AI which heavily inspired me to broaden my horizon in that area

But first, here’s a walkthrough of Day 1’s sessions—and an expanded look at the lively Team Topologies Panel and its revealing poll results.

Keynote by Simon Wardley: From Here to There and Back Again

Simon Wardley’s keynote was a masterclass in how situational awareness can transform organizational decision-making. He started by grounding us in the basics of Wardley Maps, illustrating how anchoring user needs and tracing the components that fulfill those needs can give us a map—rather than a mere diagram—of our systems. This mapping approach offers a dynamic way to visualize movement and evolution, giving teams insight into which elements should be custom-built, standardized, or outsourced. It’s far more powerful than static frameworks like SWOT, because it shows us not just what we have, but where and how it might shift over time.

What especially resonated with me was Simon’s expansion into the role of AI in decision-making. He suggested that “the future of system architecture is all about deciding who makes what decisions and where”—a perspective that fits perfectly with the research I’m doing on staffing, procurement, and make/buy/customize strategies in fast-flow environments. Pulling in references to DDD domain classifications, the Cynefin framework, and now Wardley Maps, I feel inspired to explore AI’s potential to aid—or even automate—certain architectural decisions. I’m more motivated than ever to dig deeper into this in the weeks and months ahead.

A quote form Simon Wardley

Architecture and Agility: A Shared Skillset by Eberhard Wolff

In a talk that seamlessly blended theory and practice, Eberhard Wolff tackled the interplay between architecture—often seen as a stable blueprint—and agility, which thrives on rapid iteration. Wolff argued that good architecture is never static; it evolves in tandem with the system and the people building it. Attempting to future-proof everything can be a trap, as unforeseen variables inevitably arise and must be addressed.

He stressed the necessity of distributing architectural decision-making across teams, rather than centralizing it with a single authority. Tools like Architecture Decision Records (ADRs) are helpful for capturing context, but genuine agility comes from empowering individual teams to adjust those decisions as the environment changes. This resonates strongly with modern concepts like Enabling Teams, which provide platforms and support so delivery teams can move quickly. Overall, Wolff reminded us that neither architecture nor agility is an end in itself; both exist to serve teams that continuously shape a product to meet evolving user needs.

Autonomous Teams Require Great Technical Leaders by Jakob Wolman & Gitte Klitgaard

Jakob Wolman and Gitte Klitgaard built on earlier discussions around agility and architecture, focusing specifically on team autonomy. They defined autonomy as having “decision power within a frame”—teams must clearly understand their boundaries and the organizational goals if they are to make effective decisions. This requires not just the right structures but also the right skills and mindset.

The talk delved into different leadership styles—Expert, Achiever, and Catalyst—and how each one impacts autonomy. Expert and Achiever Leaders can bring strong direction, but might overshadow team input or push overly rigid targets. Catalyst Leaders, on the other hand, foster a coaching culture that encourages experimentation and shared responsibility. The speakers underscored that autonomy isn’t a quick fix: it’s an investment in both the technical and interpersonal development of the team. When done thoughtfully, however, it yields higher engagement, better solutions, and greater resilience to change.

Gitte and Jakob on stage

Beyond Estimates (Estimates and “NoEstimates”) – Let’s Explore the Possibilities by Woody Zuill

Woody Zuill is known for pioneering Mob Programming and the NoEstimates movement, and his talk challenged many assumptions about how teams plan and track progress. He questioned the value of spending excessive time on estimates, especially when the real world often defies our predictions. Instead, Woody advocated investing that effort in collaborative development, continuous delivery, and rapid feedback loops.

His broader theme was the importance of cultivating a learning culture. By regularly inspecting and adapting, teams can pivot more effectively in response to evolving requirements. Woody was careful not to dismiss estimates entirely; rather, he emphasized balancing the need to plan with the inherent uncertainty of complex work. In the context of architecture and agility, his perspective resonated as a reminder to stay flexible and open to rethinking plans when confronted with new evidence.

Quote from Woody

4 Lenses on Organizations by Simon Rohrer

Simon Rohrer’s talk provided a rollercoaster-like bird’s-eye view of organizational structures, exploring four different lenses for understanding how companies function: (1) Team Topologies and value stream mapping, (2) nested and networked value, (3) the viable system model, and (4) the liquid organization. The key takeaway was that no single lens is sufficient; instead, these frameworks work best in unison to give a holistic picture of how value flows through an organization.

From organizing around value streams to supporting fluid, informal networks, Simon highlighted the importance of a multifaceted approach. For instance, the viable system model stresses that every part of the organization must be able to adapt to changes, while the liquid organization concept encourages cross-functional, spontaneous interactions. All four lenses, in Rohrer’s view, reinforce the central idea that people—not processes—ultimately determine how well an organization can adapt to complexity.

Simon Rohrer on stage

Team Topologies Panel: Poll Results, Challenges, and Opportunities

The first day concluded with our Team Topologies Panel, which brought together a remarkable lineup of experts: Evelyn Van Kelle, Tsvetelina Plummer, Simon Rohrer, Carola Lilienthal, Kenny Baas-Schwegler, and Eduardo Da Silva. As the moderator, I kicked off the session by sharing results from a quick Mentimeter poll, asking attendees two questions: “Team Topologies is now 5 years old—what’s your assessment so far?” and “How far along are you in adopting Team Topologies principles in your organization?” The responses indicated broad enthusiasm for the framework’s emphasis on fast flow and clear team boundaries, but also revealed a tangible gap between conceptual understanding and real-world implementation.

A significant portion of the audience reported that they had experimented with Team Topologies’ concepts—such as defining team types (stream-aligned, enabling, platform, and complicated subsystem) and using interaction modes (collaboration, X-as-a-Service, facilitating)—but found it challenging to scale these approaches in larger, more traditional environments. Some respondents specifically mentioned difficulties in drawing clear lines of ownership and responsibility, leading to confusion about who should be involved in key decisions and when. Several also noted that while they loved the theoretical foundations, stakeholder buy-in remained a critical hurdle.

Drawing on these poll insights, the panelists shared real-world stories of successes, stumbling blocks, and unexpected lessons learned. They stressed the importance of maintaining open communication channels not just within delivery teams, but also with leadership and other business functions. This helps ensure that the foundational ideas of Team Topologies—especially fast flow and team cognitive load—are understood as a holistic approach rather than an IT-centric initiative. Many organizations, the panelists observed, underestimate the cultural shift required: it’s not just about changing job titles and team structures but reevaluating how different teams interact and support each other’s goals.

Ultimately, the panel concluded that adopting Team Topologies is a journey, one that benefits from iterative experimentation, organizational empathy, and a willingness to involve multiple stakeholders early on. As Team Topologies turns five, the shared perspective was that its principles remain highly relevant, but leaders and practitioners must recognize that success hinges on balancing theory with an adaptable, context-specific strategy. The lively debate, peppered with audience questions and firsthand accounts, underscored the potential of Team Topologies to reshape how modern tech organizations think about architecture, autonomy, and flow.

The Team Topologies Panel

A Promising Start and What’s Next

Day 1 at Agile Meets Architecture offered a mix of strategic insights and practical frameworks. I came away convinced that the lines between “agile” and “architecture” are blurring in the best possible way—both disciplines demand continuous adaptation, team empowerment, and clarity of purpose. I’m excited to see where these discussions head on Day 2, particularly how organizations can effectively incorporate the evolving landscape of AI into their decision-making processes.

A final note of gratitude: it was incredibly rewarding to share this environment with friends and peers from various communities. The open, intimate setting of this conference ensures that the talks, workshops, and informal chats all reinforce one another. If Day 1 is anything to go by, Agile Meets Architecture will continue to be a beacon for those looking to unite fast-flow practices with robust architectural thinking. Stay tuned for further updates as I dive deeper into the themes and insights from this remarkable event!