How Rabobank's Team Swap initiative sparked GenAI innovation

When fourteen engineers challenged our Innovation Model

We stared at the screen, looking at the fourteen eager names who had signed up for our team swap experiment. What had seemed like a brilliant idea - having engineers temporarily join other teams to share knowledge - was now keeping us up at night.

Why? Because our first attempts had fallen flat. Engineers waited months for swaps that never happened. When they did happen, we felt the pressure to come up with tasks that matched their skills and aligned with our deliverables, which often detracted from genuine learning opportunities. Our dream of spreading GenAI innovation across Rabobank was getting buried under logistics and objectives.

But here's the thing: those fourteen names weren’t a problem. They were telling us something crucial about what engineers really want.

The Wake-Up call

Traditionally, companies approach innovation as a top-down process: executives set the vision, managers create the roadmap, and engineers execute. But is this truly the most effective way? Could this rigid structure stifle the creativity and adaptability that are essential for genuine innovation? What if the key lies in flipping the model entirely, letting ideas and innovation flow from the ground up?

At Rabobank, we were deep into cutting-edge GenAI work and knew other teams could benefit from this knowledge. Instead of launching yet another training programme or documentation hub, we decided to try something different: inviting engineers to join us for a sprint.

The response was overwhelming. Fourteen engineers signed up! But that’s when reality hit hard: just how were we going to incorporate this into our way of working?

Where we stumbled

Our first instinct was to treat these swaps like mini-projects. We tried to:

• Match engineers to specific tasks.
• Define concrete deliverables.
• Measure success through completion.

It felt like a misstep. As a team, we struggled to balance hosting swappers with our existing workload, leaving us feeling burdened rather than energized. Knowledge sharing often took a backseat to task completion, prompting us to pause the program for six months to reassess.

The lightbulb moment

Then it hit us: we were solving the wrong problem.

The real value wasn’t in completing tasks - it was in the experiments, the failures, the unexpected discoveries. Engineers weren’t signing up to be temporary workers; they wanted to be explorers.

The pivot that changed everything

We threw out the old playbook. Instead of matching skills to tasks, we started matching curiosity to experiments. Here’s what that looks like:

• Every swap is now an expedition into the unknown. We’ve just started this new approach with one engineer currently exploring Azure API Management’s GenAI features. Two more swaps are planned for January and February, where engineers will test emerging tools in real scenarios alongside their buddy.
• Success isn’t measured by completed tasks but by insights gained. Some engineers are investigating new frameworks. Others are documenting unexpected limitations. Every discovery is valuable.

Swaps are becoming more intentional and lightweight. We’re starting with a swap each month and hope to increase the frequency as we refine the process and build momentum.

The results (so far)

We’re still early in this journey, but the initial outcomes are promising:

• The swaps have facilitated meaningful cultural exchanges. Engineers are learning new tools and adopting fresh approaches, such as improved communication techniques shared by a joinee. These exchanges have brought valuable insights into our workflows and highlighted opportunities for improvement.
• The swaps are providing a window into other Areas and Tribes within Rabobank, allowing participants to understand their unique challenges and collaborate on solutions.

Engineers have gained practical knowledge, including how to efficiently build a RAG solution on Azure, which has broadened their technical perspectives and skills.

The bigger picture

Here’s what keeps us up at night now: not worry, but excitement. We’re not just changing how teams work together. We’re changing how innovation happens.

Think about it: in a world where AI and technology evolve daily, can any single team hold all the answers? Or do we need to create environments where knowledge flows freely, experiments are encouraged, and every engineer - regardless of level - can contribute to innovation?

What’s next

We’re currently running swaps on a monthly basis, with plans to increase frequency as the process matures and it becomes second nature to us. It’s ambitious. It might not work. But that’s exactly the point - we’re experimenting, learning, and growing together.

Because ultimately, that’s what innovation is really about: not just building new technology, but building new ways for people to learn and create together.

How does your team approach innovation and break down silos? Share your strategies with me and let’s learn from each other!