Document Folder Batch
Category: Healthcare Standards
Node id: document-folder-batch
Standards: PHI, IEC-62304-Class-C
Process every PDF in a local input folder: extract, run the quality/review gate, and write each result as a JSON file — auto-approved documents to output_dir, documents needing human review to review_dir. Filesystem in / JSON out; no GCS, no FHIR. Concurrency-bounded; one file's failure never aborts the batch.
What it does
- Batch PDF ingestion from a local folder
- Quality/review gate: auto-approved vs human-review routing
- One structured JSON file written per document (split across output_dir / review_dir)
- Reuses the single-document extraction core (no divergence)
Runtime
- Performance tier:
variable
Input ports
| Name | Type | Required | Description |
|---|---|---|---|
input_dir | text | No | Input directory of PDFs (overrides config; default /data/pdf) |
output_dir | text | No | Output directory for auto-approved JSON (overrides config; default /data/pdf-output) |
review_dir | text | No | Output directory for JSON needing human review (overrides config; default /data/pdf-review) |
Output ports
| Name | Type | Required | Description |
|---|---|---|---|
summary | json | No | Counts: total, passed, needs_review, failed + input/output/review dirs |
results | json | No | Per-file results (file, output path, document_type, needs_review, error) |
review_queue | json | No | Paths of JSON files written to review_dir (need human review) |
needs_review_count | number | No | Number of documents routed to human review |
success | boolean | No | True when no file failed |
Settings
Basic
| Setting | Name | Type | Required | Default | Allowed values | Description |
|---|---|---|---|---|---|---|
| Input Directory | input_dir | string | No | /data/pdf | Folder containing the PDFs | |
| Output Directory | output_dir | string | No | /data/pdf-output | Folder for auto-approved JSON | |
| Review Directory | review_dir | string | No | /data/pdf-review | Folder for JSON needing human review | |
| Suffix Filter | suffix | string | No | .pdf | Only process files ending with this (case-insensitive) | |
| Model | model | string | No | gemini-2.5-pro | Gemini model (falls back to GEMINI_MODEL) |
Auth
| Setting | Name | Type | Required | Default | Allowed values | Description |
|---|---|---|---|---|---|---|
| GCP Project | gcp_project | string | No | Vertex AI project (falls back to GEMINI_PROJECT / GOOGLE_CLOUD_PROJECT) | ||
| GCP Location | gcp_location | string | No | global | Vertex AI location (falls back to GEMINI_LOCATION) |
Advanced
| Setting | Name | Type | Required | Default | Allowed values | Description |
|---|---|---|---|---|---|---|
| Max Concurrency | max_concurrency | number | No | 4 | min 1 | Files processed in parallel |
| Min Confidence | min_confidence | number | No | 0.6 | min 0; max 1 | Below this, a document is routed to human review |
Runtime notes
- Runs on the hydra-server host filesystem — dirs are paths on the VM, not the client.
- Documents are routed to review_dir when confidence < min_confidence or the model flagged fields.
- needs_review_count + review_queue feed a downstream quality-gate / notification stage.
See Build any workflow for how to put nodes together and Ports, types and data flow for how port types connect.
Quelque chose n'est pas clair ? Envoyez vos commentaires sur la documentation par courriel.