An AI-powered design sprint agent ecosystem built for Bayer, a global life sciences company. As design consultants, we designed a connected system of three agents to democratize innovation sprints across a distributed global workforce.
Bayer's innovation practice was expensive, geographically concentrated, and dependent on a small pool of trained facilitators. Sprint knowledge evaporated after every session. We were brought in to change the underlying system.
Bayer's innovation process was fragmented across divisions, with each team operating in isolation. Sprint ownership needed to shift from individual facilitators to a shared, interconnected organizational structure.
Sprint methodology was trapped in a few regional hubs, preventing global teams from running independent sessions. Scaling required a system that could transfer facilitator knowledge without losing fidelity.
We combined semi-structured interviews, contextual inquiry, and sprint artifact analysis. Four core insight clusters emerged — and became the direct architecture of the product.
From four clusters of research findings, we came up with three "How Might We" questions.
We ideated and co-created with Bayer's stakeholders together. After 2 weeks of collaboration, we created this ecosystem with 3 key capabilities for the design agent.
We ran two co-creation workshops, 2 rounds user testing with Bayer's innovation leads, product owners, and facilitators. Concepts were rapidly prototyped and tested in real sprint contexts to validate assumptions before committing to hi-fidelity.
Three fully designed agents — Sprint Navigator, Sprint Memory Library, and Sprint Enablement Hub — brought to high fidelity and presented to Bayer's innovation leadership team.