Read Handwriting with Confidence Review

Transcribe handwritten intake forms and route uncertain blocks to staff for review.

Transcribe handwriting blocks with per-block confidence scores
Apply a confidence threshold to flag uncertain fields
Build a review queue with page numbers and text

Time to complete: ~10 minutes

Prerequisites

A SAGEA_API_KEY exported as an environment variable
A scanned intake form in PDF, PNG, JPG, TIFF, or WEBP format, max 100MB, ideally 300 DPI
Python 3.9+ with the requests package installed
Set up document processing
ARVA OCR model card

Step 1: Transcribe the intake form

Use model="arva-ocr" with JSON output to get handwriting blocks and confidence.

curl -X POST https://api.sagea.space/v1/ocr/process \
  -H "Authorization: Bearer $SAGEA_API_KEY" \
  -F model="arva-ocr" \
  -F [email protected] \
  -F output_format="json" \
  -o intake.json

Expected block shape:

{"text": "...", "pages": 1, "confidence": 0.992, "blocks": [{"type": "handwriting", "page": 1, "bbox": [x0,y0,x1,y1], "confidence": 0.87, "text": "..."}], "model_used": "arva-ocr"}

ARVA OCR supports 150+ languages and hits 99.2% on clean 300 DPI scans. Handwriting scores lower, so review matters.

Step 2: Flag low-confidence blocks

Split blocks into accepted and needs-review sets with a threshold.

curl -X POST https://api.sagea.space/v1/ocr/process \
  -H "Authorization: Bearer $SAGEA_API_KEY" \
  -F model="arva-ocr" \
  -F [email protected] \
  -F output_format="text"

Step 3: Build a review file for staff

Write a Markdown review queue that counselors can check quickly.

with open("review-queue.md", "w") as f:
    f.write("# Intake review queue\n\n")
    if not review:
        f.write("All blocks above threshold. No review needed.\n")
    for i, r in enumerate(review, 1):
        f.write(f"## Item {i} (page {r['page']}, conf {r['confidence']:.2f})\n\n")
        f.write(f"Type: {r['type']}\n\n")
        f.write(f"Text: {r['text']}\n\n")
        f.write("Staff check: [ ] confirmed / [ ] corrected\n\n")
 
print("wrote review-queue.md")

Standard pages cost $1.00 per 1000 pages, complex pages $2.50 per 1000 pages.

Verify

Error	Cause	Fix
`401 Unauthorized`	Missing or invalid API key	Export a valid key as `SAGEA_API_KEY` and retry
`400 unsupported_format`	File type not in PDF, PNG, JPG, TIFF, WEBP	Convert the scan to JPG or PDF and retry
`413 Payload Too Large`	File is over 100MB	Downsample or split the scan, then retry
`429 rate_limit_exceeded`	Too many requests at once	Back off with retries and reduce concurrency