BankingDocument Intelligence

Case study

Scaling a major remediation review with technology

A ‘Big 4’ Australian consultancy used Document Intelligence to help review 64,000 remediation cases, classify 45 million documents and produce regulator-ready evidence within a year.

Big four consultancy – Australia

Sydney - Australia
  • 45,000,000Documents ingested
  • 64,000Case reviews
  • 200,000Daily document ingest
  • 95%Classification and dating success rate

Summary

Key outcomes at a glance

  • 45 million documents ingested across 64,000 case reviews
  • 95% document classification and dating success rate
  • Average handling time reduced by 50%
  • 95%+ documents identified as irrelevant to the review outcome and excluded from reviewers’ attention
  • Recordsure technology interfaced with the consultancy’s existing workflow and MI solutions

The challenge

A big four consultancy was engaged to review 64,000 cases as part of a major remediation exercise for a large Australian retail bank. The work involved assessing more than 45 million documents and completing the exercise within a year, while maintaining a strong evidence trail for the regulator. 

The technology needed to handle more than 200,000 documents a day, integrate with existing workflow and MI solutions and scale to support around 500 users. The consultancy also targeted a 40% reduction in average handling time without compromising review quality. 

Alongside the scale and speed of the exercise, the consultancy needed a definitive record of its review findings and the evidence supporting them that could withstand regulatory scrutiny. 

Average handling time was reduced by 50%—exceeding the original target—across 64,000 case reviews and 45 million documents.

The Recordsure solution

Recordsure used its technology to create a tailored approach to analysing and reviewing the large volume of documents involved in the remediation exercise. 

The solution brought together Document Intelligence, Workflow automation and Reporting & Analytics (MI), connecting document analysis and review evidence with the consultancy’s existing workflow and MI environment. 

Analysing documents at scale

Recordsure’s Document Intelligence classified, dated and analysed documents using AI and business rules. The technology classified and dated 45 million documents across 70 file types dating back to 2006, applying bespoke categories relevant to the remediation exercise. The technology also scaled dynamically to accommodate peaks and troughs in document volumes. 

The process achieved a 95% classification and dating success rate. It also identified and surfaced relevant information from the documents, helping reviewers focus on evidence pertinent to the review rather than working through large quantities of irrelevant material. 

Connecting document analysis with the review process

Once documents had been ingested, the consultancy’s existing workflow tool was triggered to begin the case. Once documents had been ingested, the consultancy’s existing workflow tool was triggered to begin the case, connecting Recordsure’s document analysis with the wider review process. 

Reviewers could mark up documents in line with the core themes of the remediation and add bespoke narrative where required. 

Evidence for each outcome

When a case reached its final outcome, an evidence pack was created containing copies of the documents used to reach the conclusion, together with the annotations and narrative applied during the review. Evidence for individual findings could be retained as electronic data or in-document mark-up and exported as PDFs, providing a record of the evidence supporting the review findings. A metadata file was also created for MI ingestion, supporting the connection between the review process and the consultancy’s existing management information environment. 

The outcomes

The consultancy completed:

  • 64,000 case reviews
  • Involved 45 million documents
  • Document ingestion volumes of more than 200,000 documents per day capable with Recordsure technology

Document review and handling:

  • Average handling time was reduced by 50%, exceeding the consultancy’s original target of a 40% reduction
  • Document classification and dating achieved a 95% success rate
  • Around 95% of documents that were irrelevant to the review outcome could be excluded from reviewers’ attention, while key relevant information was surfaced to support the review.  

The consultancy also gained an evidence pack for each concluded case containing the marked-up documents used to support the outcome. Recordsure interfaced with the existing workflow and MI solutions, bringing document analysis, the review process and supporting evidence together within the wider remediation approach. 

Engagement snapshot

  • 2 yearsClient for
  • 520Users
  • 45,000,000Total documents ingested
  • 64,000Total case reviews

Frequently asked questions

Recordsure technology classified, dated and analysed documents using AI and business rules, identifying key data points and surfacing relevant information for reviewers. The exercise covered 45 million documents across 70 file types dating back to 2006. 

Average handling time was reduced by 50%, exceeding the consultancy’s original target of a 40% reduction. 

Document analysis surfaced key information relevant to each case, while around 95% of documents identified as irrelevant to the review outcome could be excluded from reviewers’ attention. 

Reviewers could mark up documents and add narrative during the review. At the final outcome stage, an evidence pack contained copies of the documents used to reach the conclusion, together with the associated annotations and narrative. 

Recordsure interfaced with the consultancy’s existing workflow and MI solutions. Following document ingestion, the existing workflow tool was triggered to begin the case, while a metadata file was produced for MI ingestion after the review. 

Download this case study

PDF · Free

Download Tech-enabled remediation reviews for Royal Commission finding – ‘Big 4’ Consultancy Firm – Australia (PDF, 403 KB)