Medical Record Collation (MRC) is a specialist provider of medical record collation services for the legal industry, with over 500 years of combined medical expertise spanning 70+ specialties. Their services support solicitors and barristers handling clinical negligence and personal injury cases.
Challenge
MRC needed to dramatically increase the speed and scale of their manual medical record collation process to meet rising customer demand and maintain their market-leading position. Processing thousands of pages daily was time-intensive, repetitive, and limited by staff capacity.
The team faced challenges in scaling operations without compromising accuracy. Manual collation was not only slow but also prone to errors, particularly with varied formats such as handwritten notes, typed reports, and low-quality scans. To remain competitive, MRC sought an AI/ML-driven solution that could intelligently process, classify, and paginate records at scale.
Solution
Firemind built an AI/ML-powered proof-of-concept (PoC) to automate MRC’s document processing workflow. Using Amazon Textract to extract text from handwritten and scanned medical records, and Amazon Comprehend to classify and extract key data, the system replicated and enhanced the existing manual collation process.
Custom classification models, combined with AWS Lambda and Amazon DynamoDB, enabled serverless scalability and reduced costs. The PoC validated a production-ready system capable of rapidly sorting, paginating, and indexing large volumes of diverse medical records, empowering MRC to process more cases while freeing staff for higher-value tasks.
Services used
- Amazon Textract
- AWS Lambda
- Amazon DynamoDB
- Amazon Comprehend
The Results
- 1,920% faster processing – 1,000 documents in 12.5 minutes vs. 4 hours
- 88% keyword accuracy in automated classification
- 80% AWS cost reduction through optimised modelling
- 500% increase in document handling capacity
- Proprietary AI/ML IP positioning MRC as a technology leader
