Celerity as Client Zero with IBM Bob

AI had moved quickly from experiment to everyday use across Celerity. Engineering, managed services and operational teams were working with a growing mix of AI assistants, each bringing potential productivity gains and a new set of questions.

Without a common way of working, AI usage risked becoming inconsistent. Development practices could fragment across teams, access to company information could go unchecked, and activity could become difficult to audit. Governance policies varied from one provider to the next, leaving open questions around intellectual property and data security that a coding assistant alone could never answer.

The challenge was no longer how to give teams access to AI. It was how to make AI a governed, repeatable part of how Celerity designs, builds and operates technology, before recommending that same approach to clients.

Celerity does not believe customers should be the testing ground for new technology, so we took that role ourselves. We started with our own engineering, managed services and operational teams, running internal Bobathons and workshops before rolling IBM Bob out into everyday engineering work, beginning with the IBM Power environments (IBM i and AIX) where we hold deep expertise and where many of our customers face real modernisation challenges.

These platforms tend to carry decades of business logic, heavily customised applications and tightly coupled processes running critical workloads that cannot simply be replaced. Putting IBM Bob to work here, on our own estate, gave us a genuine test of what governed AI adoption looks like at enterprise scale, and surfaced lessons that go well beyond the technology itself, covering governance, adoption, engineering standards, security and organisational change.

Those lessons shaped the platform we now bring to customers.