Project / Case study
Saveika
Saveika helps field workers capture household visit signals, calculate mental health risk, and give supervisors a clearer view of households that may need urgent follow-up.
01 / Research question
What it asks
How can community health workers collect household signals and help supervisors identify people who may need mental-health follow-up?
02 / Experimental design
Scope
The platform provides guided household-visit intake, risk scoring through Gemini with deterministic fallback logic, supervisor review tools, English and Nepali support, seeded demo accounts, and installable PWA support.
03 / Evaluation
How it measures
Saveika combines Supabase Auth, Postgres with row-level security, a mobile-first interface, and map views built with Leaflet and OpenStreetMap. The repository also includes tests with Vitest and Testing Library.