Send a PDF or URL — get structured JSON back in seconds. Runs on your infrastructure. Zero cloud data exposure.
Vendor name, date, total, currency, and line items — extracted from any PDF or web page, returned as clean JSON.
Key obligations, parties, and red flags surfaced automatically. Know what you're signing before you sign it.
Category, urgency, and suggested action — route high-priority messages automatically without reading every inbox.
Processing runs locally via Ollama. No document ever leaves your server. GDPR-safe by architecture.
POST a job, GET the result. Webhook callbacks supported. JSON in, JSON out. Integrate in an afternoon.
Every result includes a confidence score. Low-confidence extractions are flagged for manual review automatically.
One plan. Cancel anytime. First month is a trial.
# 1. Submit a document
curl -X POST https://api.inksky.net/backoffice/jobs \
-H "X-API-Key: your_key" \
-H "Content-Type: application/json" \
-d '{"type": "invoice_extraction", "input_url": "https://yoursite.com/invoice.pdf"}'
# → {"job_id": "abc123", "status": "queued"}
# 2. Get the result (usually ready in <30s)
curl https://api.inksky.net/backoffice/jobs/abc123/result \
-H "X-API-Key: your_key"
# → {
# "vendor_name": "Acme GmbH",
# "invoice_date": "2026-03-01",
# "total_amount": "1240.00",
# "currency": "EUR",
# "line_items": [...],
# "confidence": 0.95,
# "needs_review": false
# }Questions? Not sure if this fits your workflow?
contact@inksky.net