Osiar

Features
How It Works
Use Cases
Pricing

Document Intelligence Platform

Extract structured data from any document using AI. Upload an image, get JSON back. Simple API, powerful results.
Learn More
# Step 1: Upload & quality check
curl -X POST https://api.osiar.my.id/v1/upload \
  -u "pk_live_xxx:sk_live_xxx" \
  -F "file=@invoice.jpg" \
  -F "template_id=019abc..."

# → { "file_ids": ["019756a2..."], "status": "ready", "quality_score": 0.92 }

# Step 2: Extract
curl -X POST https://api.osiar.my.id/v1/extract \
  -u "pk_live_xxx:sk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{ "template_id": "019abc...", "file_ids": ["019756a2..."] }'

// Response:
{
  "data": {
    "nik": "3201234567890001",
    "name": "JOHN DOE",
    "birth_date": "01-01-1990"
  },
  "confidence": 0.95
}

Everything You Need

Template-Based Extraction

Define what fields to extract. Our AI handles the rest. Support for text, numbers, dates, currency, and arrays.

Smart Rule Engine

Extraction rules match documents by keywords and extract data via regex — no AI cost. LLM only as fallback.

Workflows

Auto-detect document type and route to the right template. One API endpoint for all your document types.

AI-Powered Template Creation

Upload a sample document, AI suggests fields. Refine with natural language. Test before saving.

Auto-Learning

System analyzes LLM results and suggests extraction rules. Over time, fewer AI calls, lower costs.

Multi-Tenant & Secure

API key auth with rate limiting, IP whitelist, monthly quotas. Each tenant isolated with their own templates.

How It Works

1
OCR
Document image is processed with AWS Textract to extract raw text.
2
Rules Match
Extraction rules try to match keywords and extract fields via regex. If successful — done, no AI cost.
3
AI Fallback
If rules fail, OCR text is sent to LLM for intelligent extraction. Structured JSON returned.

Use Cases

ID Cards (KTP, SIM, Passport)
Extract NIK, name, address, birth date from identity documents.
Invoices & Receipts
Pull vendor name, amounts, line items, tax from any invoice format.
Utility Bills
Extract meter numbers, usage, billing period from PLN, PDAM, Telkom bills.
Medical Records
Parse patient info, diagnoses, prescriptions from hospital documents.
Financial Documents
Extract data from bank statements, tax forms, insurance claims.
Custom Documents
Any document with structured data. Define your own template and extract.

Simple Credit-Based Pricing

Pay per extraction. No monthly fees. Buy credits, use them when you need.

Starter

1,000

credits~500 extractions
Best for testing

Growth

10,000

credits~5,000 extractions
For production use

Enterprise

Custom

volumeUnlimited extractions
Dedicated support

Developer-Friendly API

RESTful API with Basic Auth. SDKs for JavaScript, Python, Go. Webhook support.
Upload Image
POST /v1/upload
Content-Type: multipart/form-data

image: (file)
template_id: "019abc..."

→ { "file_ids": ["..."], "status": "ready" }
Extract Document
POST /v1/extract
{
  "template_id": "019abc...",
  "file_ids": ["019756a2..."]
}
Execute Workflow
POST /v1/workflows/:id/execute
{
  "file_ids": ["019756a2..."],
  "route": "pln"
}

Ready to Extract?

Start extracting structured data from your documents in minutes.
© 2026 Osiar. Document Intelligence Platform.