From manual pulls to one risk story.

At the start, the reporting ran on a weekly cycle of manual work: pull the data, clean it up, and republish the dashboards, by hand, every week.

For Issue Management, that cycle took two days. By the time a report was published, the data was already getting older, and it kept aging until the next week's pull. Every hour spent pulling and republishing was an hour that couldn't go into improving the data or the questions we were asking of it.

So the first job was clear: take the people out of the copy-and-paste loop.

I replaced the weekly pulls with automated ones, built with Power Query, Reporting Services, and API integrations.

Issue Management reporting went from a two-day manual weekly process to a fully automated platform that refreshes every hour. Going from weekly to hourly was a huge win: the numbers now reflect the last hour, not last week.

I redesigned Remediation reporting the same way, with automation and analytics that give near real-time visibility into remediation work, so leadership can follow progress and spot emerging risks as they happen.

Just as important, the hours that used to go into pulling and republishing came back to the team. Everything after this stage was built with that time.

Once the data refreshed on its own, I could start watching its quality all the time, not just once a week.

So I built hygiene dashboards that monitor the quality of the data entered into Optro and Azure DevOps (ADO). They flag missing and invalid fields, like a record without an owner or a date that doesn't make sense, so the team can find and fix the problem quickly, at the source.

Behind them are data governance and quality controls across Issue Management and Remediation, with standardized field validation and monitoring. The result is more accurate, more consistent, and more reliable reporting.

With the data automated and checked, I consolidated it into a shared semantic model.

Alongside it, I built and centralized a portfolio of 35+ Power BI dashboards across Internal Controls, Risk, Remediation, and Policy Exceptions. Having the data and the dashboards in one place means people start from the same numbers instead of their own copies.

The model also made the next step possible: AI-enabled data agents in Microsoft Copilot, layered on top. Instead of waiting for a new report, people can ask a question in plain language and explore the risk and control data themselves.

I'll be honest: Copilot isn't perfect. It can misread a question, and it's not a replacement for a well-built dashboard. But as an exploration tool it's great. It gets people to a first answer fast, gives stakeholders self-service access to risk and control insights, and reduces reliance on manual reporting and analysis.

The time we no longer spend pulling and republishing now goes into the biggest step yet: bringing everything together in the Risk & Control Hub, built on Snowflake.

The work is standardizing data across platforms, so issues, remediation, controls, and policy exceptions line up and can be read side by side instead of in separate silos.

The 35+ dashboards and the governance behind them are the foundation for the move to the Enterprise Risk Management and Risk & Control Hub operating model. Soon that foundation will let us tell a holistic ERM story, one view of risk across the enterprise, instead of a handful of separate ones.

• Timeliness: Issue Management reporting that refreshes every hour instead of once a week, and near real-time visibility into remediation.
• Trust: standardized validation and hygiene dashboards that catch bad or missing data where it's entered.
• Consistency: 35+ dashboards across four programs, centralized in one place.
• Self-service: Copilot data agents that let people explore the data without waiting on a manual report.
• Foundation: the data groundwork for the Enterprise Risk Management and Risk & Control Hub operating model.

See my full experience · ← All projects