I help teams make possible organizational change tangible: co-designing and prototyping how roles, decisions, interactions, experiences and work could function with AI, so people can explore, test and learn before committing to scale.
Organizations often know something needs to change, but the future state remains abstract: a strategy slide, a target picture, a list of use cases. It is hard to commit to, and even harder for people to imagine working in.
Meanwhile, automating individual tasks inside an unchanged system can simply move friction elsewhere. The step gets faster, and review, coordination and exception handling around it grow to compensate.
I help teams co-design and prototype possible organizational changes so people can explore, test, and learn before committing to scale.
How we work
See the system around the challenge: who is involved, how work and decisions flow, and where change would matter.
Develop possible interventions together with the people who will live with them.
Make the change tangible enough to experience, whether a role, a decision logic, an interaction or a workflow.
Try it, gather feedback, refine, and decide what deserves to scale.
Depending on the challenge, we may prototype any combination of these. Deciding what humans, AI and automation should each contribute is one useful design consideration among them, not the whole design.
Typical format: 1–2 days
Bring diverse teams together around meaningful organizational challenges, explore the system, create possible interventions, and develop tangible prototypes or experiments.
Not necessarily a coding hackathon. Prototypes can be new roles, decision logic, experiences or ways of working, not only software.
Typical format: 2–3 weeks, or a 5-day intensive
Take one important challenge deeper through co-design, prototyping, testing, and iteration.
Automation asks how a step can be done without a person. Redesign asks what the system should look like now that a new capability exists. It can remove steps, move decisions, change roles or make something possible that was never attempted.
It depends on the context. AI is strong at synthesis, drafting and pattern recognition. Humans remain essential where consequences are significant, context is unwritten, trust is built, and someone must be accountable. Prototyping makes these choices testable rather than theoretical.
Treat an agent as a participant with defined scope, visible actions, a person who owns its outcomes, and a clear escalation path.
Design for it deliberately: keep people in positions where they choose rather than only approve, make disagreeing with a system output legitimate, and preserve enough hands-on practice that expertise does not erode.
People react to something concrete instead of debating a target picture.
Assumptions are tested before significant investment, not after.
Those who co-design the change are more likely to carry it.
If you have a challenge that is ready to be made tangible, I am glad to think through whether a hackathon or a design sprint would serve it best.
Or write directly to sara@bermudez-t.com.