Deadlines, incidents, complex systems and constant context switching. For many developers, working under pressure is a daily reality. At the same time, we live in a world where information, notifications and expectations constantly compete for our attention. In one word: stimulus overload. The question is therefore not how to avoid pressure, but how to manage it effectively and sustainably.

In this interactive workshop, we explore what happens when tension and stimuli build up and how you can continue to perform precisely when they do.

Talent Manager and former elite athlete Wouter Stoltz shares his personal story of performing at the highest level, dealing with setbacks and reinventing himself after a serious injury ended his sporting career. The lessons from elite sport prove surprisingly applicable to the daily reality of software development: maintaining focus under pressure, recovering after exertion and managing tension effectively.

We also examine the science behind stress, attention and performance. What happens in your brain when pressure rises? Why can thinking become more difficult precisely when you need to be sharp? And how does an overload of stimuli affect our concentration, energy and decision-making?

Inspired by insights from neuroscience and the growing focus on reducing stimuli and recovery, participants discover why sustainable performance is not only about working harder, but also about managing attention and energy more intelligently.

The workshop combines personal stories, scientific insights and immediately applicable techniques. Participants learn to recognise their own stress signals, gain practical tools for regaining focus at peak moments and develop a personal plan for high-pressure situations.

This is not a session about reducing workload. It addresses a far more relevant question: how do you ensure that pressure, complexity and stimuli work for you rather than against you?

The AI Circle is an invite-only breakout for managers with an AI focus. If you are not yet part of the community but would like to find out whether you qualify to participate, please contact events@teamrockstars.nl

What happens to your data platform if a foreign government demands access to your storage? How does your hyperscaler's unpredictable pricing strategy affect your quarterly budget? What happens to your enterprise architecture if a critical vendor technology is no longer supported? What are the legal and GDPR risks when your data passes through third parties or foreign subprocessors? In this session, we explore the urgency of sovereignty and portability for your data platform. Based on your organisation's needs and requirements, we outline the technical consequences. After all, complete sovereignty and portability also have disadvantages…

Using the Data Platform Navigator, a framework developed by Team Rockstars IT, we show architects, engineers and programme managers how to find the right balance between cloud convenience and absolute control over their own data, step by step.

What will you take away from this session? Clear questions and a practical framework for determining your own balance between cloud convenience and sovereignty.

High performance does not come from making everyone work in the same way. It comes from understanding how different people think, communicate, focus, make decisions and handle pressure. In a neurodiverse organisation, this is even more important. The challenge is not to erase the differences, but to turn them into a strength.

In this session, we examine how to build a high-performing team in which different minds can bring out the best in one another. Drawing on practical experience with teams in complex environments, we share what helps and what proves counterproductive when people have different needs around structure, communication, autonomy, focus, feedback and energy.

We cover psychological safety, clear expectations, working agreements, decision-making, a sustainable pace, conflict, ownership and trust. The goal is not to create a perfect team in which everyone works in the same way. The goal is a team whose members understand one another better, make use of each other's strengths and deliver meaningful results without burning out.

In this interactive breakout, technology philosopher and AI analyst Wietse Hage explores how AI systems are evolving and the opportunities and challenges this creates for the future of work. He covers the most important developments, limitations and trade-offs, as well as the impact this technology has on how we build and use software. As co-host of the AI Report podcast and newsletter with Alexander Klöpping, Wietse follows developments closely and translates them into clear insights.

Following a brief introduction by Wietse, this session becomes an open conversation with engineers. Guided by questions from the group, we explore the possibilities and limitations of AI in practice: where are the boundaries of current models, how much control do you still have as an engineer, and how do you deal with uncertainty, bias and emergent behaviour?

We also examine the impact on the profession. As more code is generated partly or entirely by AI, the engineer's role increasingly shifts from builder to orchestrator. How do you direct AI systems effectively, safeguard quality and make considered technical and ethical decisions? What does this shift mean for knowledge work, and how do we prepare for a future in which collaborating with AI becomes a natural part of the job?

An in-depth session for engineers who want to understand how their role is changing in an era in which AI is playing an increasingly prominent part in software development and knowledge work.

This session demonstrates why cloud sovereignty is no longer a theoretical issue for organisations such as VGZ, but a strategic necessity, and which risks arise when you fail to act. Drawing on the approach, initiatives and results of VGZ and Team Rockstars IT, participants gain practical insights into achieving control over cloud sovereignty through European cloud choices, portability and a realistic transition strategy.

AI is rapidly becoming the new term for digital transformation. But the real question is no longer whether organisations should do something with AI; it is how to deploy AI at scale and connect it to measurable business value.

In this session, I take participants from using isolated AI tools to thinking like an Agent Boss: someone who not only uses AI agents, but also directs and organises them and connects them to real business outcomes.

Topics include:

• why AI is about more than productivity or automation;
• how to move from experiments to AI at scale;
• how to connect AI to strategic value, customer impact and new ways of working;
• what this means for leadership, teams and the role of technology professionals.

Many organisations invest in talent development, knowledge sharing and collaboration, while also finding that engagement can no longer be taken for granted. Hybrid working, individualisation and a growing need for autonomy make it more difficult to sustain connection, ownership and knowledge exchange.

Yet these factors are crucial to the success of teams. When professionals continuously learn from one another, actively share knowledge and feel connected to their colleagues and the organisation, teams become stronger, performance improves and service delivery becomes more consistent.

Community Manager Sjoerd Kuijten (Team Rockstars IT) discusses this with Chapter Lead Software Development Bart Sanders (Alliander). Do not expect a theoretical presentation, but an open conversation about how Bart Sanders has successfully put this into practice at Alliander over the past few years. Join this roundtable session and share your thoughts and questions during the breakout.

Choose this breakout for a guided tour of the Johan Cruiff Arena. At the start of the tour, please report to the meeting point at the event venue near the information desk.

Learn how to build frontier Agents leveraging the latest open-weight frontier models. Know and identify the right patterns to circumvent LLM shortcomings and fight the right balance between accuracy, costs, safety and data privacy.

It is becoming increasingly clear that AI is the defining technological revolution of this century. It is fundamentally changing how we work, learn and live together. However much we might wish otherwise, we cannot address the societal questions that follow using existing logic or familiar solutions. As legislation, frameworks and institutions fall short, we must build a new ethical infrastructure to turn all this technological change into genuine progress.

Against this backdrop, we reconsider the role and value of human beings. Behind all the prominent discussions about jobs, data centres and Big Tech lies a much deeper shift in the role AI plays in our lives. A shift in which we are gradually handing over not only our work, but also our social and personal lives to these human-like machines. The human story as we know it has reached a turning point. What will our next chapter look like?

A developer opens their laptop in the morning, gives an AI agent a prompt and has a working feature before lunch. The code compiles, the generated tests pass and the pipeline is green. Yet nobody dares to release it.

With Agentic AI, we have a tool that generates code at a speed no developer can match. But development speed was never the real problem. The bottlenecks have been the same for years: unclear requirements, unstable test environments, insufficiently representative test data, slow feedback loops and architectures that treat testability as an afterthought. AI solves none of these issues. It only makes their consequences arrive sooner and hit harder.

The obvious response is to have AI generate the tests as well. But when both the implementation and the tests are generated from the same prompt, you are no longer validating software. You are validating an AI's interpretation. A passing test only proves that the code does what the test expects, not what the user actually needs. This dilemma is central to the session: when can you trust AI-generated tests, and when do they merely create an illusion of quality?

Drawing on practical experience in large enterprise environments, I discuss the choices teams must make before AI tooling can genuinely add value. What requirements must your test environments and test data meet? When is mocking justified, and when does it hide a structural problem? And how do you convince an organisation focused primarily on building new features that quality thinking starts when discussing a requirement, not after the PR has been merged?

This is not an argument against AI. It is an argument for Quality Engineering as a discipline, rather than QA as a phase at the end of the software delivery process.

Chaos in the SOC: technology, communication and responsibility under pressure

A suspicious login from abroad. Shortly afterwards, a process execution alert. An L1 analyst wondering whether it is serious enough to escalate. An L2 immediately asking for more logs. A client who mainly wants to know whether they will still be able to work tomorrow. And somewhere in a Teams call, fifteen people are now silently staring at one another.

Welcome to the SOC.

In this session, I take the audience through a recognisable, practice-oriented account of how security teams operate under pressure. No theoretical “best practices”, but a realistic look at what happens from the moment an alert arrives until technology, communication and responsibilities begin to overlap.

Using a fictional case grounded firmly in practice, we follow an incident through the different layers of the SOC: from L1 to L2 and potentially on to L3 and management. Along the way, technical analysis, human error, stress, tunnel vision, escalations and communication challenges converge. These moments reveal that security is not only about tooling or detections, but above all about collaboration, clear responsibilities and decision-making under pressure.

The session includes recognisable situations for anyone working in security operations, incident response or technical operations organisations. In an accessible, narrative style, I show how small signals can develop into major incidents, how teams respond as pressure rises and why tabletop exercises are essential preparation.

What happens when an IT company decides to put its name on the Ajax shirt?

In this session, Robin Hogendorf and Kimmy Gommers take you behind the scenes of the partnership between Team Rockstars IT and AFC Ajax. No polished sponsorship pitch, but an honest account of making bold choices.

How do you ensure that such a partnership becomes more than a logo beside the pitch? How do you build visibility and trust in an extremely competitive market? And how do you translate that into concrete impact on recruitment, business and thought leadership?

With a transparent, data-driven look at strategy, activations and results, we share the journey we have taken with Ajax over the past two years. From skybox to hackathons, from brand awareness to a multimillion-euro business. How do you continue to make decisions under pressure that are not only exciting, but genuinely strengthen your brand?

Expect an open account of having the courage to choose, continuing to build and sometimes simply taking the leap without guarantees.

How VodafoneZiggo makes a complex network transformation scalable

How do you translate a strategic ambition into automation that actually works within the network?

VodafoneZiggo faces the challenge of enabling ever-higher internet speeds on its existing coaxial-fibre network. To achieve this, approximately 15,000 Remote PHY Devices (RPDs) must be rolled out over the coming years. Manual provisioning is simply not fast, scalable or reliable enough.

In this talk, Solution Architect Oscar van der Leij and Product Owner Aidan Langelaan share their joint journey from strategy to a working Network Automation platform. Over the course of more than eighteen months, they worked with IT teams, network specialists and engineers to build a solution that largely automates the configuration and provisioning of an RPD after installation.

But automation proved to be about far more than technology alone. Requirements only became truly clear during testing with engineers, specialist knowledge was scarce and dependencies such as network connectivity regularly affected progress. At the same time, existing processes had to be simplified and different domains had to learn to work closely together.

A practical and transparent account of brownfield integration, orchestration, iterative development and the lessons learned along the way, and above all why successful automation only emerges when technology, processes and people come together.

This talk will be in English, limited space (10p)

When interacting with technology, most of it is done via a visual medium, yet for us humans, speech is the more native form of communication. With the AI evolution picking up speed, should we abandon our “old” ways of developing and look into designing applications and the way we interact with them from scratch?

What would a world look like when every user has their own UI customized to their needs and preferences, and what would that mean for our work? Will we still have frontend developers? How will it change the way we create API’s and handle information provision when the UX and UI are always different?

Join me in this interactive thought experiment and discover what our work might look like in the future, now that we have AI not only as a coding companion, but also as a user interface.

Choose this breakout for a guided tour of the Johan Cruiff Arena. At the start of the tour, please report to the meeting point at the event venue near the information desk.

Join us at the bar for an hour and discover how much influence you have over your own happiness at work. This interactive Happy Hour session gives you practical insights and tools to gain more energy, enjoyment and impact from your work.

We show why happiness at work and performance go hand in hand, how to stay in control of your work when things get tough and which small actions can make an immediate difference for both you and your colleagues.

You will leave with a personal workplace happiness check, practical tools for tomorrow and, of course, a surprising dose of Rockstars confetti (be there to find out ;)).

In this interactive panel from KPN Digital Experience, you will hear different perspectives on how we encouraged developers, including the sceptics, to explore AI tools and learn how to use them. Rather than making AI use mandatory, we sparked curiosity by creating room to experiment, learn out loud and discover firsthand what actually works within a real codebase.

We share our lessons, practical experiences and an honest view of what truly made a difference and what did not. Team Rockstars IT joins the conversation, allowing us to share experiences and insights from both KPN and Team Rockstars.

Ask your questions, gain new insights and be inspired.

Forty-five minutes of insights, inspiration and honest conversations about turning scepticism into curiosity.

Europe is not merely replacing gas turbines with wind turbines; it is building an entirely new digital backbone for how energy is generated, balanced, traded and billed. Every choice made along the way, from cloud provider to data format, quietly determines how much control we retain.

In this session, we first zoom out to examine the forces radically changing European energy policy and market structures. We draw a parallel between decades of dependence on oil and gas and the new dependency risk hidden in cloud contracts and relationships with AI vendors. We then zoom back in on what this means in practice for the platforms and security approaches you manage today: how to design for an exit using open table formats such as Apache Iceberg, portable compute on Kubernetes and open protocols instead of closed APIs. We also discuss why resilience and openness must be explicit architectural requirements.

A session for engineers and technology leaders who want to build quickly without inadvertently surrendering their future flexibility.

This talk will be held in English. Limited space (10p)

This talk introduces an agentic purple-teaming environment: a multi-agent IDE designed for secure, reproducible offensive security research. The platform combines coordinated AI agents with strong sandboxing, execution isolation, and deterministic replay to enable safe experimentation, collaborative pentesting workflows, and auditable attack simulation.

The session explores how agentic systems change security tooling, how reproducibility and containment become critical when autonomous agents interact with targets and tools, and what this means for the future of AI-assisted cybersecurity. Attendees will see practical lessons and potential pitfalls from building an environment where AI agents can coordinate reconnaissance, exploitation, analysis, and reporting while remaining observable, constrained, and safe to evaluate.

This talk will be in English

The moment your organization wants to become less dependent on American tech, the bar suddenly looks impossibly high. GitHub, Cloudflare, npm, your base images: there is always another thread running to the US. And because we can't do it perfectly, we end up doing nothing. That is the Purity Trap.

Antoine lowers that bar. No large organization is switching over completely any time soon, and it doesn't have to. Not every team can move to a European cloud today, but every team can become less dependent today.

With lessons from his own work, from the Dutch national COVID platform to Volt Europa and Enexis, he shows how to take those first steps without blowing up your roadmap. The approach in outline: know your crown jewels, close the tap, and keep the lights on. You'll walk out with a concrete list you can start on tomorrow. And with the realization that "not perfect" still counts.

AI agents can write code, invoke tools and move through a codebase at incredible speed. But speed alone does not produce reliable software. In real software development, generated code is not yet delivered work. A change only becomes valuable once it has passed through validation, review, acceptance and delivery.

This session starts with a simple observation: onboarding a new developer into a serious software project takes time. That developer must build domain knowledge, understand the architecture, learn the testing strategy, follow team conventions, know the security rules and understand how work moves from idea to production. Why, then, do we expect an AI agent to perform the same work reliably based on a single vague prompt?

The problem is not only that models hallucinate. The larger problem is that much of engineering is implicit. Intent lives in conversations. Rules live in people's heads. Validation lives in habits. Context is scattered across tickets, documentation, tests, API contracts, architecture decisions and delivery pipelines. Agents make those weaknesses painfully visible.

In this talk, we explore the Engineering Triad as a practical model for agentic development:

• Prompt Engineering makes intent explicit.
• Context Engineering makes knowledge explicit.
• Harness Engineering makes execution explicit.

Using a concrete password-reset example, we show how a simple request such as "add password reset" actually contains many hidden engineering decisions: security rules, token policies, API contracts, email infrastructure, validation, logging, test scenarios and human acceptance. We translate that vague request into explicit artefacts such as a SPEC.md, BDD-style scenarios, AGENTS.md rules, scoped custom agents, reusable skills and validation gates.

The key message: agentic development does not replace engineering discipline. It makes engineering discipline even more important. A good environment for agents is also a good environment for developers.

The budget for your AI use case has been secured, the business case has been approved and the initial results are promising. But then comes the most difficult question: how do you actually determine whether an AI solution is successful?

In this session, I show how we approach this at Rabobank. We define what “success” means early on and capture it in evaluation suites (evals). This not only makes the value of AI solutions measurable, but also helps us build an organisation-wide evaluation strategy that gives business, risk and compliance teams insight into performance and risks.

Using practical examples from the FinTech world, I show how to create evals, use them to compare models objectively and deploy them to continuously monitor the quality of AI solutions. Expect an interactive session with live demos, real-world examples and room to exchange experiences.

This breakout session features two consecutive TED Talks. Registering for this session gives you access to both talks.

This session brings together two distinctive perspectives on a highly relevant theme: performing under pressure. In a TED-style setting, neurobiologist Brankele Frank and former elite goalkeeper and executive Edwin van der Sar explore the impact of stress on both the brain and daily functioning. Drawing on neurobiology, Brankele explains how our brains respond to pressure, what prolonged stress does to us and why recovery is at least as important as performance.

Edwin van der Sar then discusses his own experiences of performing at the highest level. How do you handle the expectations of millions of supporters, the pressure of decisive matches and the responsibility of leadership? Drawing on his career as an international player, his time at leading clubs such as Ajax and Manchester United and his years as Ajax's CEO, he shares personal lessons about focus, resilience and making decisions under intense pressure.

Together, Brankele and Edwin show that successful performance is not only about knowledge, talent or technical skills, but also about mental strength, self-awareness and the ability to keep making the right decisions under pressure.

For generations, we’ve designed organizations around efficiency, processes and technology, and then taught humans how to fit inside them. Now AI is accelerating that logic beyond anything we’ve seen before. Faster decisions. Faster output. Faster change. And once again, the message to humans is simple: adapt.

But what if we’re optimizing the wrong side of the equation?

We already know that humans perform at their best when the right conditions exist around them. Culture, trust, purpose, autonomy and the freedom to think differently are not soft extras. They shape what becomes possible.

AI gives us an extraordinary opportunity. Not simply to reinvent what humans do, but to rethink the environments in which humans do it.

We keep asking what humans must become to survive the future of work. Maybe the future of work should start by asking what it means to be human.

Choose this breakout for a guided tour of the Johan Cruiff Arena. At the start of the tour, please report to the meeting point at the event venue near the information desk.