Signature story · from Scaling cloud and engineering capability
Teaching an enterprise to build differently
Scaling cloud and engineering capability at Fidelity
The environment
Fidelity between 2016 and 2020: a global asset manager with a digital transformation strategy on paper, deep on-premise habits in practice, and cloud experience concentrated in a handful of enthusiasts.
The constraint
The strategy assumed thousands of engineers could design event-driven, cloud-native systems. They could not, yet. Sending people to certification courses produced badges, not systems, and every project team relearned the same lessons at full price.
The intervention
I stopped treating capability as a training problem and treated it as an engineering problem. In APS we codified reusable infrastructure as code so projects inherited a working foundation, and ran live-coding workshops where engineers built real serverless systems in the room. In Personal Investing I scaled the idea into a learning dojo: self-contained katas using dynamic reference architectures, each one walking through a full solution including its pipeline, designed so teams in different global locations could upskill concurrently. Alongside it, my team built the event-driven platform those skills would be used on, and I ran Event Storming sessions so the business defined the domain boundaries with us.
The result
Tentative AWS experiments became a multi-year cloud strategy with live cloud-native applications. The dojo was adopted by multiple teams across global sites. Reusable infrastructure cut the cost and time of every subsequent project, and the event-driven platform became the backbone of the digital experience ecosystem.
What I learned
Capability scales when it ships as a product: versioned, self-service, opinionated and tested. A reference architecture someone can run teaches more than a slide deck someone must believe.
Evidence and artefacts
- Role detail, Director of Software Architecture, APS
- Role detail, Director of Architecture, Personal Investing