At AI Incubation, we operate as a 0 to 1 product lab inside IBM. We identify emerging AI capabilities early and turn them into validated, scalable enterprise applications. Our team works at the intersection of research, rapid prototyping, and product strategy. We build new product categories rather than incrementally improving existing ones. From natural language to application coding platforms to computer using agents, we explore how AI reshapes how enterprise teams build, reason, and operate. We believe there is no better moment to be designing new products!
We are a lean team optimised for speed, ownership, and craft.
Builders first. Everyone prototypes. Everyone experiments. Everyone ships.
Action over approval. We prioritise momentum and real world validation over consensus.
Low ego, high standards. We care deeply about quality and how we collaborate.
Research to product. We translate early signals into working systems quickly.
Everyone codes. Not everyone is a software engineer, but everyone engages directly with the product surface.
This created a core product tension: build for engineers who want production-ready code, or build for everyone else who just wants a fast draft? Our first instinct was to optimise for power and polish. That was wrong.
Our team operates like a founder collective inside a global enterprise. We are comfortable with ambiguity and allergic to unnecessary process. We move between research papers and shipping deadlines with equal fluency. We bridge what is emerging with what is useful. We stay hands on with frontier capabilities while maintaining the urgency and accountability of an early stage startup.
Every incubation effort is supported by a dedicated internal community of IBMers. We leverage IBM’s global network to run rapid feedback loops, conduct user testing, and validate design decisions at scale.Our communities are not passive focus groups. They are active collaborators. We ship frequently, publish changelogs, and evolve products in public. This creates early adopters and power users before a product reaches the wider market. Even our most experimental initiatives are grounded in real workflows and real user needs.
As part of our incubation work, I had rare access to the IBM Library and Archive. Working alongside an exceptional team of librarians, we studied decades of IBM’s cultural history to surface the company’s tone of voice, visual language, and typographic identity.The goal wasn’t nostalgia, it was clarity. We wanted to understand the wit, discipline, and design confidence that shaped IBM’s character, and reinterpret it for a new generation of AI products.
