Konverter: AI Document Intelligence for Legal Publishing

Konverter is an AI-powered document intelligence system designed to support the digital transformation of complex, unstructured PDF publications. Many legal and government organisations still rely on PDFs that are difficult to publish as accessible web content, reuse as structured data, or integrate with emerging AI systems.

Konverter helps by transforming PDF documents into accessible, responsive HTML and structured JSON-LD which makes information easier for both people and machines to use. The system primarily uses Docling, VLMs and OCR-based extraction to identify document content, semantic structure and metadata. A confidence-based, human-in-the-loop review process allows uncertain results to be validated before publication This improves reliability while keeping people in control.



Its modular architecture also provides opportunities for future applications such as AI training datasets, retrieval-augmented generation (RAG), chatbots, agentic workflows and automated document processing. While initially developed for complex legal publications, Konverter is designed to be adaptable to other industries and specialised document workflows.