The question we keep hearing
One of the most common questions we get from customers is about scaling: “We have AI use cases in production. What’s the next move? How do we really build the next version of our enterprise using AI?”
The mistake: treating AI in silos
The most common mistake we see is treating AI tools and use cases in silos. A lot of people are using AI in their daily work today, but does it really move the needle when so much of the underlying work remains unchanged? Most of that remaining work is meetings, handovers, explaining and aligning. The real opportunity isn’t just making individual tasks faster. It’s reducing the friction between them.
The shift: a new way of working
The key shift is not adopting a new AI tool. It’s adopting a new way of working.
Instead of knowledge living in meetings, chats and individual documents, we create shared, structured knowledge that becomes the team’s shared memory, accessible to both people and AI. AI can then work with the same context throughout the entire lifecycle – from idea and requirements, through development and testing.
What this looks like in practice
– Treating the repository as the team’s shared memory
– Using AI as a partner across the entire delivery lifecycle, not only for coding
– Moving knowledge from meetings, chats and documents into reusable, machine-readable assets
– Creating traceability through specifications, decision logs and repository-based documentation
– Continuously validating quality through AI-assisted reviews and coverage checks
Where to start
We recommend focusing on building this shared memory, establishing specification-driven ways of working, and using AI consistently to improve quality and reduce knowledge loss.
This is one of the field lessons behind our AI Factory offering – helping enterprises move AI pilots into real, compounding business impact.
Are we spending too much time making individual tasks faster with AI, and too little time rethinking how work actually gets done?
