Statistics and causal thinking
Regression, survival analysis, quasi-experimental methods, forecasting, and applied biostatistics for public-interest questions.
Skills & Expertise
Evidence-based capability across research, analytics engineering, and AI-ready systems
Skills
What I can build, run, and improve.
This page is intentionally specific. I care less about broad claims and more about whether a skill has been used to solve a real delivery, quality, or decision problem.
Core capability areas
Data architecture and operations
ETL design, validation checks, reproducible reporting, warehouse thinking, and cross-team data handoffs that survive real operations.
AI and ML systems
Feature design, model framing, monitoring, and responsible deployment patterns grounded in statistical discipline rather than hype.
Field data systems
Survey design, digital tool building, enumerator training, and high-frequency checks across low-connectivity environments.
Dashboards and evidence translation
Operational dashboards, reporting automation, and concise outputs for ministries, funders, programme teams, and executives.
Capacity building
Practical training in R, Stata, reproducible workflows, and data quality habits for teams that need more than one-off analysis help.
Tools and methods
Analytics and statistical methods
RStataPythonSPSSSurvival analysisForecastingRegressionCausal inference
Data engineering and infrastructure
SQLPostgreSQLETL designValidation checksData QAGit workflowsReproducible reporting
AI and machine learning
XGBoostClassification workflowsFeature engineeringModel monitoringLLM prompt designResponsible AI framing
Field systems and digital data collection
KoboToolboxODKREDCapSurveyCTOCommCareEnumerator trainingHigh-frequency checks
Communication and dashboarding
Power BIR ShinyQuartoR MarkdownPresentation decksPolicy briefsExecutive reporting
Evidence of use
| Area | Evidence |
|---|---|
| Data architecture | ETL and quality workflows for 1,000+ household studies and partner dashboards |
| Field systems | Managed tool design, training, and QA across multi-country studies with 50+ field staff |
| Statistics | Mixed-methods and quantitative analysis for health systems, oncology, and impact evaluation work |
| Dashboards | Built Power BI, Quarto, and Shiny outputs for monitoring, reporting, and policy dialogue |
| AI transition | Created portfolio projects in predictive ML, LLM-assisted evidence briefing, and dashboard-driven alerting |
Current edge
The direction I am sharpening now is the overlap between rigorous research data work and modern data systems:
- reproducible ML and model monitoring
- AI applications that support evidence translation
- data engineering habits that make analytics easier to trust and scale