Automating digital evidence processing for NSW prosecutors with AI-powered workflows

What they were up against

Prosecutors faced growing volumes of digital evidence with slow, manual processes: inconsistent file formats and naming, time-consuming sorting and classification, limited automated extraction or summarisation, and difficulty identifying confronting content early—creating risk to both case efficiency and staff wellbeing.

What we did

• Designed and built a secure evidence ingestion pipeline on Microsoft Azure with Windows Defender scanning
• Implemented automated audit logging and processing workflows to ensure integrity and compliance
• Delivered OCR and debundling capabilities to handle diverse file formats
• Built intelligent file renaming and classification using Azure Cognitive Services
• Integrated AI and LLM-based summarisation to accelerate evidence review
• Developed sensitive-content detection to reduce unnecessary staff exposure to confronting material
• Integrated with ODPP systems including the Brief Service Portal and matter repositories
• Delivered in phased enhancements to progressively improve capability

What changed

A secure, phased evidence processing pipeline was delivered with automated scanning, audit logging, OCR, intelligent renaming, AI summarisation, and sensitive-content detection—making evidence faster and easier to review while preserving integrity, compliance, and staff wellbeing.