Kaide-LABS/Sand-FDE-App

Two data-quality-first artifacts for a Sand Technologies FDE application: a Katsina State health-data trust audit, and a Borno State PHC worker reallocation optimizer # Sand Technologies FDE Application Two unsolicited technical artifacts built to accompany a Forward Deployed Engineer (Nigeria) application to Sand Technologies. Sand's own public position is "Foundations Over Features": that generative AI and dashboards built on top of broken or ungrounded data are the wrong first move, and that the unglamorous work -- finding out where data can't be trusted, then acting on data once it can be -- is the right one. Both artifacts try to demonstrate that kind of work directly, in Sand's own stack (Python, PostgreSQL, Airflow, dbt, Superset), rather than describe it. They are deliberately sequential, not two unrelated demos: - **Artifact One** answers the question that has to come first: *can this data be trusted at all?* It is a deterministic data-quality audit of Katsina State's health reporting, and it finds real, citable problems in the data itself (reporting attrition, biologically impossible coverage values) without touching what any of it means for care delivery. - **Artifact Two** assumes data that has cleared that bar and asks the next question: *given trustworthy data, what physical, real-world action does it justify?* It targets Borno State's primary-healthcare workforce maldistribution and produces a specific, checkable staff-reallocation proposal -- zero-sum, zero new hires -- rather than another dashboard. Read that as the throughline: trust the data first, then use it to make something happen. **Live reports:** Artifact One · Artifact Two -- static snapshots of each pipeline's generated output, hosted via GitHub Pages so a reviewer can see the result without running anything. Regenerated by hand from a real pipeline run, not hand-edited; the code in each artifact directory below is what actually produced them. ## Artifacts | | What it does | Status | |---|---|---| | `artifact-one-data-quality/` (report) | Ranks Katsina State's 34 LGAs by how untrustworthy their reported health data is, and says exactly why -- re …