Inveeze Admin Guide
This comprehensive guide is designed for System Administrators and Operations Managers using the Inveeze SaaS platform. It covers user management, detailed pipeline configuration, billing, and troubleshooting.
1. User Management & Access Control
Inveeze uses a role-based access control (RBAC) system scoped to each Tenant.
Roles
- Admin: Full access to all settings, billing, pipelines, and user management.
- Finance Manager: Can view billing, invoices, and usage reports. Cannot modify pipelines.
- Auditor: Read-only access to documents and audit logs.
- Member: Standard access to upload documents and view results.
inviting Users
- Navigate to Settings > Team.
- Click Invite Member.
- Enter the email address and select the Role.
- The user will receive an email to join your specific Tenant.
2. Pipeline Configuration
Pipelines are the core of Inveeze. They define how documents are processed, from ingestion to export. Configuration is managed via the Visual Editor or directly via JSON Configuration.
Structure
A pipeline config is a JSON object defining a list of nodes and swimlanes.
{
"nodes": [
{
"id": "classify_1",
"type": "classification",
"settings": { ... }
}
],
"swimlanes": [
{
"id": "ingestion",
"label": "Ingestion"
}
]
}
Example: Invoice Processing Pipeline
Use this JSON template for a standard Accounts Payable workflow.
{
"name": "Standard AP Workflow",
"nodes": [
{
"id": "class_main",
"name": "Document Classification",
"type": "classification",
"settings": {
"provider": "google_gemini",
"model": "gemini-1.5-flash",
"categories": ["invoice", "receipt", "contract", "unknown"]
}
},
{
"id": "extract_inv",
"name": "Invoice Extraction",
"type": "extraction",
"settings": {
"schema": {
"invoice_number": "string",
"date": "date",
"total_amount": "currency",
"vendor_name": "string",
"line_items": "table"
},
"validation_rules": [
{
"field": "confidence",
"condition": ">",
"value": 0.85
}
]
}
},
{
"id": "export_erp",
"name": "ERP Webhook",
"type": "webhook",
"settings": {
"url": "https://api.your-erp.com/invoices",
"method": "POST",
"headers": {
"Authorization": "Bearer {{secrets.ERP_TOKEN}}"
}
}
}
],
"swimlanes": [
{
"id": "ingestion",
"label": "Ingestion"
},
{
"id": "processing",
"label": "Processing"
},
{
"id": "review",
"label": "Review"
},
{
"id": "export",
"label": "Export"
}
]
}
Validator Configuration
You can add specific validation rules to extraction stages.
| Rule Type | Description | Example |
|-----------|-------------|---------|
| required| Field must not be null | {"field": "total_amount", "rule": "required"} |
| regex | Must match pattern | {"field": "invoice_number", "regex": "^INV-\\d+"} |
3. Node Configuration Details
Each node in the nodes array requires specific properties.
Common Properties
id(string): Unique identifier (e.g., "node_1").name(string): Display name in the editor.type(string): Node type (classification,extraction,webhook,logic).layout(object): Position{ x: 250, y: 100 }(managed by editor).
1. Classification Node
Determines the document type.
{
"type": "classification",
"settings": {
"provider": "google_gemini", // or "openai", "azure"
"model": "gemini-1.5-flash",
"categories": ["invoice", "receipt", "contract"]
}
}
2. Extraction Node
Extracts structured data based on a schema.
{
"type": "extraction",
"settings": {
"schema": {
"field_name": "type" // string, date, currency, table
},
"validation_rules": [
{ "field": "total", "condition": ">", "value": 0 }
]
}
}
3. Webhook Node
Sends data to external systems.
{
"type": "webhook",
"settings": {
"url": "https://api.example.com",
"method": "POST",
"headers": { "X-API-Key": "{{secrets.KEY}}" },
"body": "{{document.data}}" // Optional custom payload
}
}
4. Tutorial: Creating a Classification Pipeline
This guide walks you through creating a pipeline that accepts an uploaded file and classifies it (e.g., Invoice vs. Bank Statement).
1. Define the Pipeline
Create a new pipeline or edit an existing one. Use the following JSON structure to set up a classification node.
{
"name": "Incoming Mail Classification",
"nodes": [
{
"id": "classify_doc",
"name": "Classify Document",
"type": "classification",
"settings": {
"provider": "google_gemini",
"model": "gemini-1.5-flash",
"categories": ["invoice", "bank_statement", "contract", "other"]
}
}
],
"swimlanes": [
{ "id": "ingest", "label": "Ingestion" }
]
}
2. Upload a File
There are two ways to feed this pipeline:
Option A: Dashboard Upload
- Go to Documents.
- Click Upload Document.
- Select your file (PDF/Image).
- Choose "Incoming Mail Classification" from the pipeline dropdown.
Option B: API Upload
Use the POST /documents endpoint with the pipelineId.
3. Review Results
Once processed, the document status will update.
- If the AI is confident, it will tag the document (e.g.,
classification: invoice). - You can view this metadata in the Document Detail view.
5. Billing & Credits
Inveeze operates on a credit-based consumption model.
Credit Usage
- Classification: 1 Credit per page.
- Extraction (Standard): 5 Credits per page.
- Extraction (Premium): 10 Credits per page (includes line-item reasoning).
Monitoring Costs
To view real-time usage:
- Go to Billing.
- Consult the Usage Record table.
- Filter by
Pipelineto identify high-cost workflows.
4. Troubleshooting & Support
Common Issues
Issue: Document Stuck in "Processing"
- Check the Pipeline Editor for valid JSON syntax.
- Verify your webhook URLs are reachable.
- If using bespoke models, ensure API keys are set in
Settings > Integrations.
Issue: Login Failure
- If SSO is enabled, contact your IT admin to verify the Inveeze Enterprise app is assigned to your user.
API & Webhooks
For developers integrating with Inveeze, see our API Documentation.
For critical incidents, email support@inveeze.com with your Tenant ID: {{tenant.id}}