Back to Portfolio

Will Ingestion AI

A sophisticated automated document processing system designed to ingest, parse, and extract critical information from legal will documents. The system leverages Azure's suite of AI services to handle unstructured document data, automatically identifying key entities, clauses, and beneficiary information with high accuracy.

Will Ingestion AI

What It Does

The project automates the traditionally manual process of reviewing and cataloging will documents. Users can upload will files through a scalable cloud infrastructure, which then:

  • Automatically extracts text and structured data from documents
  • Identifies key legal entities and provisions (executors, beneficiaries, assets, conditions)
  • Processes documents in parallel for high throughput
  • Stores extracted information in a structured format for downstream analysis and retrieval
  • Provides intelligent insights powered by generative AI

Results & Impact

The system successfully automates a previously time-intensive manual process, reducing document review time from hours to minutes while maintaining legal accuracy standards. The combination of deterministic extraction (Document Intelligence) with generative AI intelligence (Azure OpenAI) provides both reliability and semantic understanding, making it a robust solution for legal document automation.

Challenges & Solutions

Document Variability & Format Handling- Wills come in numerous formats (PDFs, scans, different legal templates) with varying quality and layouts. Ensuring reliable extraction across this spectrum required careful tuning of document intelligence models and fallback strategies. Accuracy and Compliance - Legal documents demand high accuracy since errors could have real-world consequences. Balancing automation speed with the need for confidence scores and human review mechanisms was critical. Scalability and Cost Optimization - Processing potentially thousands of documents requires an efficient, cost-effective pipeline. Managing concurrent requests while staying within Azure service quotas required careful architecture decisions around batching and throttling. Data Privacy and Security - Handling sensitive legal documents containing personal information necessitated secure end-to-end processing with proper encryption, access controls, and compliance considerations.

Technologies Used

Azure Blob StorageAzure Durable FunctionsAzure Document IntelligenceAzure Cognitive ServicesAzure OpenAIPython

Project Type

AI Development

Date Completed

19/01/2026

Client

Legal Client