Case studyAnalytics Professional
Analytics in healthcare — Fundación CTIC (2025–today)
My current role: the analytics a healthcare institution uses to measure and decide, and the strategy it uses to govern its artificial intelligence under UNE-ISO/IEC 42001:2025.
- Analytics
- Data governance
By the numbers
42
analytics products in use in Power BI
20
leaders supported, across 15 processes
12
AI opportunities identified
23
instruments of the AI management system
Context
In healthcare, administrative and clinical decisions depend on institutional data that historically lived scattered and without clear governance.
The challenge
Bring data governance, quality and analytics to the decisions of the Foundation's administrative and clinical leaders.
How I did it
Process analytics for two audiences
I develop analysis and visualization models in Power BI for administrative and clinical leaders: 42 products, including 23 control dashboards. An indicator keeps a single institutional definition, even if each audience sees it with its own level of detail.
Data governance and quality in healthcare
I structure institutional data quality with rules for completeness, duplicates, reconciliation and thresholds, under Colombia's habeas data regime and Law 1581. Personal data is used for its authorized purpose, not for anything that can technically be reached.
Dashboards by process, not by area
I organize dashboards by process because an area can show good indicators while the process crossing it piles up waits and rework. I design the improvement plans; the owners of each process carry them out.
The institutional AI strategy
I lead the AI strategy under UNE-ISO/IEC 42001:2025: 12 opportunities identified, 7 cases formally evaluated and 3 prioritized. Not every problem needs AI, and not every AI initiative needs an agent.
A management system that can be demonstrated
We are building 23 instruments of the management system —policies, procedures, matrices and criteria—, 8 of them finished. To review them I built the ISO 42001 expert agent I publish in the showcase: it cites the standard or declares the gap, and no ruling moves forward without a person.
Impact
- 42 analytical products in use for 20 leaders across 15 processes and about 75 users.
- About 60% less effort preparing information; 10 analysis plans under follow-up.
- 12 AI opportunities identified, 7 use cases evaluated, 3 prioritized and 2 documented.
- Institutional data quality as a practice, not an event.
My purpose is not to get the organization to use more artificial intelligence, but to help it use the right artificial intelligence for relevant problems.