How to build the university that didn’t exist: Lessons from CODE University

What changes when a university is designed around student agency?

In this episode, Manuel Dolderer, Founding President of CODE University of Applied Sciences, shares how personal missions, project-based learning, open spaces, faculty coaching, and co-creation can prepare students for an unpredictable future. Listen to the full episode or read the summary of our conversation below:

Also available on:

Summary

Universities have traditionally prepared students by deciding which knowledge and methods they will need later. That model becomes harder to defend when future roles are difficult to predict and artificial intelligence can perform many of the structured tasks that dominate conventional teaching and assessment. Manuel Dolderer, Co-founder and President of CODE University of Applied Sciences in Berlin, argues that higher education should place greater emphasis on students’ capacity to choose, act, and learn through meaningful challenges.

CODE was founded in 2016 to create that kind of environment. Its model connects personal purpose with project-based learning, an open campus, accessible faculty, qualitative feedback, and student participation in developing the institution itself. The result offers a broader lesson for university leaders. Student agency cannot be added as a graduate attribute while the curriculum continues to reward compliance. Learners need repeated opportunities to experience themselves as people who can make decisions and create something that did not exist before.

Artificial intelligence exposes a deeper educational problem

Manuel sees artificial intelligence as a source of pressure on teaching and assessment practices that were already losing relevance. Universities often ask students to take in information, apply an expected method, and produce a predetermined output. These are increasingly the kinds of tasks AI can support or complete, while institutions cannot reliably detect every use of the technology.

The challenge therefore reaches beyond introducing an AI policy or adding digital tools. Universities need to reconsider what meaningful learning looks like when repetitive intellectual work can be delegated. Manuel hopes this pressure will encourage institutions to move away from practices that survived largely because they were familiar and scalable.

CODE’s model has strengthened his belief that motivation begins with purpose. Students who see no connection between an assignment and something they value have a clear incentive to use AI as a shortcut. A student who genuinely wants to understand a problem approaches the technology differently because the objective is learning rather than task completion.

If you want to understand something, you won’t use AI to figure it out for you because you are deeply curious about it.

This does not remove the importance of professional knowledge. Domain expertise remains necessary, but it is insufficient preparation for uncertain work and social environments. Students also need confidence that they can interpret unfamiliar situations, take responsibility, and have an effect on the world around them.

Agency has to be designed into the learning experience

CODE begins by asking students why they are at university and what they hope to accomplish. Wanting a degree does not provide an adequate answer. Students are encouraged to develop a personal mission and use it to guide choices about what, how, and why they learn.

You have to have some sort of personal mission.

This emphasis responds to a pattern Manuel associates with traditional education. Students follow a prescribed route and become skilled at satisfying other people’s expectations. The danger appears after graduation, when they are expected to show initiative and shape a future they have had little opportunity to practise influencing.

At CODE, project-based learning gives agency a practical form. The first semester provides orientation because students may need time to adapt after many years in conventional education. From the second semester onwards, they propose or choose projects within digital product development. Professors help them connect each idea to academic learning goals.

You’re learning because they’re related to the project. They’re related to something you want to achieve, you want to understand, you’re curious about.

Knowledge gains an immediate context. Students encounter concepts and methods because their projects require them, while meaningful problems supply a reason to persist. Choice is paired with responsibility because learners must define what they want to achieve, work with others, and make decisions as the project develops.

The campus and faculty role follow the pedagogy

Learning design also shaped CODE’s physical environment. While seeking investment, the founders met the creator of Factory, a Berlin co-working community. He was preparing to open a second site and saw a close fit between the proposed university and the collaborative environment he wanted to build. CODE’s first 88 students joined the co-working space in 2017.

The arrangement revealed what project-based learning needed from a campus. Large numbers of classrooms were less useful than smaller rooms, informal corners, and shared areas where teams could work for extended periods. Students could use the environment around the clock for projects, study, events, and social activity. The campus became part of the community rather than a container for scheduled teaching.

Faculty visibility supports the same idea. Professors work in an open shared area and remain approachable when they are on campus. Their role centres on mentoring, coaching, and guiding rather than primarily transmitting content. They help students frame projects, connect experience to learning goals, and reflect on development.

That role can be difficult because faculty cannot rely fully on the habits formed during their own academic careers. Progress may be hard to judge from one day to the next, and meaningful learning develops over time. CODE introduced supervision circles in which faculty could discuss moments of uncertainty and success, share experience, and learn from one another.

You have to actually figure this out for yourself.

Supporting that uncertainty became an important form of leadership. It allowed faculty to develop a new professional identity together while the learning model was still evolving.

Students helped build the institution they joined

Starting from scratch gave CODE a strong organising idea and attracted people who were already convinced by the approach. Early faculty and professional staff largely selected themselves into an institution whose methods were experimental and whose details were still unresolved.

The founders extended that openness to students. The first two intakes were explicitly invited to act as co-creators, contributing feedback and helping redesign elements of the learning concept as practical experience revealed what worked and what remained unclear.

You have to help build this university.

This approach reduced the need for traditional direction from a single leader. A core team of around 10 to 15 people worked closely, discussed failures and successes, and treated the university as something they were learning how to build together. The strong founding idea created coherence, while participation supplied the information needed to improve it.

Student co-creation also embodied the agency the curriculum sought to develop. Learners were trusted to influence their environment rather than simply receive it. Their feedback could lead to changes because the institution regarded their experience as evidence and their participation as part of the work.

Established universities face more embedded structures and cultures, but the principle can still inform reform. Leaders can involve students in curricular development, create forums for faculty to examine uncertainty, and treat implementation as an iterative learning process. The central requirement is to give participation a real connection to decisions.

Feedback can develop agency and make the model scalable

CODE defined a pioneer’s mindset and skillset to describe what graduates should develop alongside professional and academic knowledge. These qualities concern students’ self-image, capabilities, and ability to thrive in uncertain settings. The university chose not to reduce them to quantitative measures.

Instead, students identify how a project might develop particular skills or aspects of their mindset. Professors and project mentors then offer qualitative feedback in personal and team sessions, based on how students act during the work. Formal grades remain linked mainly to professional and academic learning goals where the institution is required to provide them.

This model demands significant faculty attention. Manuel believes greater use of structured peer feedback could make it more efficient and extend its relevance to larger institutions. Students would contribute to one another’s learning while practising how to give useful feedback, a capability with value beyond university.

The wider aim is to let students experience themselves as capable of taking on consequential problems. Personal agency cannot be postponed until graduation or expected to appear after years of following prescribed instructions.

We have to allow students to develop that while they’re studying.

For Manuel Dolderer, the university needed for an uncertain future begins with what students repeatedly experience. When learners can pursue meaningful projects, influence their institution, receive developmental feedback, and take responsibility for choices, agency becomes part of their education rather than an aspiration attached to its end.

Ready for more?

As artificial intelligence reshapes how knowledge is accessed and applied, universities face a growing challenge: how to help students develop the curiosity, judgement, and agency needed to thrive in uncertain environments. This requires moving beyond content delivery and creating learning experiences that encourage students to take ownership of their development.

• Building a generation of changemakers: The UCE Model explores how the University of Canterbury is creating an entrepreneurial learning environment centred on mindset, confidence, belonging, and experiential learning. Like CODE University’s approach, it highlights the importance of helping students see themselves as capable of creating change and acting on opportunities.
• Demystifying entrepreneurship and entrepreneurship education examines entrepreneurship as a broader competence rather than simply a pathway to business creation. The article argues that universities should help learners develop creativity, initiative, opportunity recognition, and problem-solving capabilities across disciplines, themes that closely align with Manuel Dolderer’s emphasis on agency and purpose-driven learning.

Stay tuned for the next episode on this series and don’t forget to follow us on your preferred podcast platform!