CARE Usage Tracker

Track CARE ecosystem adoption across public health dashboards. A scraper mirrors upstream metrics for tracked partners into dated JSONL snapshots, a normalizer derives compact JSON artifacts, and a static Next.js site (built with care·ui) presents them.

Live site: https://usage.ohc.network

Trackers

Tracker	Source	Metrics
ABDM	NHA ABDM dashboard	ABHA numbers created, health records linked, ABHAs linked, Scan & Share tokens, daily/weekly trends, statewise splits

Only data for partners listed in config/<tracker>/partners.yaml is collected and published — nothing else is scraped or aggregated.

How it works

trackers/abdm/scrape.ts      →  data/abdm/raw/<date>/*.jsonl   (dated snapshot + manifest)
trackers/abdm/build-data.ts  →  data/abdm/normalized/*.json    (summary, partners, trends, meta)
site/ (Next.js static export)→  site/out/                      (reads normalized JSON at build)

A GitHub Actions workflow (scrape.yml) runs the pipeline daily at 02:00 IST and commits the snapshot; Cloudflare Pages rebuilds the site on push.

Development

Requires Node 22+.

# pipeline (repo root)
npm ci
npm run scrape:abdm                      # scrape today's snapshot
npm run scrape:abdm -- --date 2026-07-09 # or a specific date
npm run build-data:abdm                  # normalize latest snapshot

# site
cd site
npm ci
npm run dev     # dev server (syncs data first)
npm run build   # static export to site/out/

Adding a partner

Append the partner's exact upstream name to config/abdm/partners.yaml. The next scrape picks it up.

Adding a tracker

1. Create trackers/<name>/ with a scrape.ts and build-data.ts following trackers/abdm/, plus config/<name>/partners.yaml.
2. Add scrape:<name> / build-data:<name> scripts to the root package.json and a job to the scrape workflow.
3. Register it in site/src/lib/trackers.ts and add routes under site/src/app/<name>/.

Data notes

• Raw snapshots are append-only under data/<tracker>/raw/<date>/; each JSONL line is { fetched_at, endpoint, params, row } with a manifest.json summarizing every request.
• Upstream daily series are sparse (only dates with activity), so derived window metrics use date arithmetic rather than last-N points.
• This is an unofficial, read-only mirror of publicly accessible dashboard data, refreshed once a day.

License

MIT