Project Scope

Clinisync exists to solve one narrow problem: clinical teams routinely receive spreadsheet exports that mix patient identifiers, demographics, and outcomes in inconsistent shapes, and they need them in a clean, de-identified, per-sheet CSV form before any analysis.

In scope

• Multi-sheet .xlsx / .xls / .ods ingestion (uploads ≤ 5 MB).
• Deterministic per-upload anonymization of NHI values to sequential IDs that align across sheets.
• DOB → Age conversion with format-tolerant date parsing.
• Removal of obvious PHI columns (Contact, Address) and empty / unnamed columns.
• Row-level dedup after transformation.
• Per-upload manifest with sheet-by-sheet stats: rows processed, duplicates removed, invalid DOB count, missing NHI count.
• Optional defenses on the upload path: API-key auth, per-IP rate limit, virus scan.

Out of scope

• Long-term storage of patient data. Cleaned CSVs land in a local csvs/ directory and are swept by a TTL sweeper (CLINISYNC_CSV_TTL_HOURS, default 24 h). The intended pattern is "download immediately, then forget."
• HIPAA / GDPR / GxP certification. The service follows privacy-conscious defaults but does not implement audit logging, access roles, encryption at rest, or BAA-grade controls. Use it inside a controlled environment, not as your compliance layer.
• Free-text NLP, schema inference, or row-level validation rules. Cleaning is rule-based and intentionally small.
• A graphical workflow builder. The cleaning pipeline is fixed in code; new rules are PRs, not configuration.

Running it

git clone https://github.com/DevilsDev/csv_procedure.git
cd csv_procedure
npm install
npm run setup           # creates uploads/, csvs/, and test fixtures
npm test                # 30 tests; all passing on main
npm run dev             # starts on http://localhost:3000

For deployment knobs (CLINISYNC_API_KEY, CLINISYNC_CSV_TTL_HOURS, REDIS_URL, CLAMAV_TCP_HOST) see .env.example in the repo.