You are hired as a Senior Project Manager at an insurer a century old, where the delivery process fights itself, agile teams running under waterfall governance, and delays that nobody owns.

An insurer cannot move fast by breaking things, hence every change has to survive risk, compliance, and the actuaries.

I built a hybrid delivery framework both sides could live with, then ran it until it stuck. I later became program manager for New Business and Retention, working across a $23M portfolio. I am now in the Corporate PMO leading three Workday Finance workstreams.

You are the newest coordinator on the floor when two pilot projects arrive for prospective clients, RBC at $7M and Desjardins at $2.5M. Losing them is not an option anyone says out loud.

Nobody hands authority to the new person, it has to be earned faster than the org chart moves.

I ran both pilots and both clients signed. Then I kept taking the work nobody owned: second in command during leadership absences, six months running client services at 100% of KPIs, and a PMO built from scratch with 65 people in it.

Your employer is acquired mid flight while five client companies hold contracts worth $77M over seven years, and every one of them is deciding whether to stay.

A contract on paper does not keep itself alive, it survives on the people delivering it staying steady through the noise.

I kept delivery running through the transition and made the retention case client by client. In parallel I stood up data mining centers in Egypt and the Philippines to expand what the company could take on.

You have 310 stores that need their modems and firewalls replaced, some of them north of the 60th parallel, and a marketing team drowning in a process that touches 184K SKUs.

A store that goes dark loses money by the hour, so there is no maintenance window big enough to hide in.

I ran both projects at once with a team of 25: the national hardware rollout store by store, and an internal app that rebuilt the merchandising process around the 184K SKUs.

You have spent a year building a certification engine for month end close, then one morning Anthropic releases finance templates that overlap your product bet.

A bet is only worth holding if you would place it again today. That morning, I would not have.

I retired the bet the same week rather than letting it fade quietly. The architecture, the enforcement pattern and the Truth Gate, became the foundation for everything I have built since.

Nobody hired you for this one, however the question would not leave. Banks and insurers will adopt AI last, and they will pay the most for controls they can enforce. Who builds the controls?

A model can be persuaded, hence whatever holds the line cannot be the model.

I directed three working systems into existence the way I ran programs, constraints first, proof before claims. The AI writes the code under my direction. The judgment is mine.