About

Before AI governance, I learned governance from data reliability.

Early in my career, I worked close to the system layer: SQL validation, ETL testing, EDW testing, batch-flow analysis, and migration logic in core banking environments. That trained me to look beyond what appears correct on the screen and ask: did the data move correctly, transform correctly, reconcile correctly, and remain reliable through the full flow?

I create operating rhythm across complex transformation.

In large-scale banking modernization work, I coordinated complex delivery across countries, vendors, components, and dependent teams. My contribution was not only tracking delivery. I helped create the governance rhythm: ownership rules, integration cadence, escalation paths, defect triage, and dependency visibility so multiple workstreams could move with less ambiguity and lower late-stage risk.

Context: Regional banking transformation, regulated banking environment

I translate cloud ambition into decision-ready governance.

In cloud migration governance work, I helped turn a large strategic ambition into decision-ready inputs: portfolio assumptions, readiness signals, security constraints, compliance considerations, budget trade-offs, and migration roadmap logic. That work reinforced a principle I now apply to AI: scaling technology safely requires governance before acceleration.

Context: Enterprise CCoE readiness, banking cloud migration governance

Transformation is workflow redesign, not just technology delivery.

In digital lending transformation, I worked across product, business, IT, security, and compliance to redesign a document-heavy lending journey into a more practical digital workflow. The lesson was clear: transformation succeeds when the workflow, controls, user experience, and operating model move together.

Context: Major Thai bank, regulated lending environment

I now test on myself what I design for enterprise work.

My current AI portfolio work is a lightweight AI workflow and knowledge orchestration experiment. It is not a production enterprise platform. It is a practical learning artifact designed to test how AI-assisted work can become traceable, reviewable, governable, and less cognitively expensive.

The flow is simple: task intake → classification → routing → execution support → visibility → human review → decision log.

The learning is bigger than the tools: AI adoption depends on workflow, governance, reviewability, operating structure, and human accountability — not only the model.