Doing more with Less - Tooling & Automation

Massena Consulting was approached by a client looking to extract text from tens of thousands of checks stored in PDF files and automatically process them in bulk. The client had previously digitized these checks and inputted the information manually, but the process was time-consuming, expensive, and error-prone. With a tight deadline, the client enlisted Massena to use machine learning (ML) to tackle this problem, requiring a high accuracy rate to meet various regulatory constraints.

The Execution

Massena quickly organized the relevant stakeholders and systematically mapped the manual process. The team recommended a custom Optical Character Recognition (OCR) solution designed to achieve at least a 95% accuracy rate. After comparing several open-source libraries and popular cloud offerings, Massena developed a solution that automatically extracted text and data from scanned documents, going beyond simple OCR to also identify the contents of fields in forms and information stored in tables.

The solution first converted each PDF document to a JPEG formatted picture, enhanced the quality of the images, and then fed them into the OCR model. The custom solution allowed for a high level of accuracy, effectively eliminating the need for manual input. The results were highly successful, with a 99.2% accuracy rate across all documentation. The solution was deployed by incorporating an automated transfer of data extracted from the OCR model directly into the client's check processing portal, automating the process end-to-end.

The solution significantly reduced work time and expense while increasing overall efficiency, resulting in:

250 Monthly Labor Hours Saved - 1.5 FTE
Estimated Savings in processing fees, time, and efficiency gains exceeding $175,000 per year
$1.9 Million in Revenue Growth
Efficiency gains allowed for the re-purposing of existing staff and greater scale.