American Water, the largest publicly traded water utility in the US, kicked off a five-year enterprise digital transformation program to modernize metering analytics, billing, customer experience, and operational systems. Accenture was the implementation partner. The program spanned seven workstreams, 40+ team members across internal and consulting staff, and a 659-line multi-year program plan. When I joined, PMO infrastructure was minimal: no shared Jira instance, no standardized cadence, and reporting relied on hand-built decks rotated through Accenture each week.
Senior leadership had visibility through Accenture's curated narrative, but no real-time, source-of-truth view of program health. Workstreams operated in silos. Risks, decisions, and action items lived in scattered meeting notes. The cost of manual coordination was compounding as the program scaled.
I built the PMO operating layer from the ground up. I stood up Jira and Confluence with role-based access, custom workflows, JQL dashboards per workstream, and parent-child epic/task/deliverable hierarchies mapped to the 659-line MS Project plan. I drove adoption through structured onboarding sessions for PMO and engineering leads, hitting 47% adoption across 41 users in 60 days.
To replace the weekly reporting cycle, I built a live Power BI executive dashboard using DAX and Power Query on Jira data, giving leadership real-time KPI visibility and cutting executive reporting prep by 60%. I facilitated sprint planning, backlog grooming, and retrospectives across all seven workstreams, lifting sprint efficiency by 45%.
To eliminate the manual coordination drag, I shipped an AI agent end-to-end that processed meeting recordings into action items, decisions, and RAID entries, then auto-exported structured updates from Excel into Jira, saving leadership roughly four hours per week. I anchored the change layer in ADKAR and managed cross-functional accountability through a RACI matrix covering 40+ resources across seven workstreams.
Measured. Documented. Repeatable.
"Governance is a product. Adoption requires the same care as a customer-facing launch: clear value, simple onboarding, real feedback loops. The AI agent worked because it solved a real, repetitive cost, not because it was AI.