The AI Document Processing Module in elDoc is designed for end-to-end intelligent document processing and automation with Human-in-the-Loop (HITL).
No traditional model training or document templates are required. elDoc can automatically understand document content, classify documents, and extract relevant data using AI. Administrators simply define which data needs to be extracted for each document type, together with the applicable processing and validation rules.
Different document types can have their own Recognition Rules and extraction requirements. Administrators can also assign a different LLM model to each document-processing use case through Recognition Rules → AI Data Extraction Model. This allows, for example, KYC documents, invoices, contracts, or claims to use different LLMs according to their processing requirements.
When confidence levels or configured business rules require additional verification, documents can be automatically routed for human validation. Based on the identified document classification, elDoc can then trigger the appropriate review and approval workflow, routing documents to the relevant users or teams.
This enables end-to-end automation:
Document Upload → AI Classification → AI Data Extraction → Post-Processing (Business Rules) → Validation & Human-in-the-Loop (if required) → Review & Approval Workflow → Final Processing → Storage
For more details on configuring AI Document Processing, Recognition Rules, and automated workflows, refer to the corresponding section of this guide.