Case Study
How a Global Management Consulting Firm Optimized Document Indexing
TABLE OF CONTENTS
The Customer
A company within the management consulting industry approached the Roots' team for help indexing and uploading documentation. This Customer’s work product is knowledge. As a result, they have a knowledge base hosted externally where they store their work product, such as presentations, white papers, research, and other written and visual content. This knowledge base allows their team to quickly find information that might be pertinent to other customers.
This Customer is a large management consulting firm with an estimated $9 billion in revenue, supported by 11,000 employees.
The Situation
As a management consulting firm, our Customer produces massive amounts of knowledge. The Customer’s knowledge base stores their products – presentations, white papers, research and other content. Because the knowledge base is massive, the team needed to classify and index all documents and content for easier navigation and management. Each document needed to be read, classified and uploaded.
Prior to partnering with Roots, the Customer was sending this work to their offshore team in India to review, classify, add and index this content into their knowledge base.
The Challenge
The offshore team classifying these documents had different coding standards, which often led to mislabeled or misclassified documents. Additionally, each document took around 15 minutes to process, making this a very time-consuming endeavor. Poor accuracy and mistakes in manual classification impacted the on-shore team’s ability to quickly find pertinent information or resources.
The Solution
Using Document Vision and Machine Learning, we trained an AI Agent to read documents, classify data according to metatags and upload into the knowledge base under the right folder. We used Natural Language Processing (NLP) and Optical Character Recognition (OCR) to enable the AI Agent to complete these tasks. Then, we enabled an easy search and find function, which not only optimized the document indexing process, but also improved the workload of the FTEs searching for these documents.
The Customer is now able to process documents in 15 seconds, compared to 15 minutes with their previous manual process. This was a +98% reduction in document processing times. Additionally, accuracy skyrocketed, alongside an 87% reduction in reliance on manual entry.
Within 8 months, the Customer reached a 650% return on investment. This extreme return is due to several factors, including the +98% decrease in time spent manually processing documents, as well as the 87% reduction in reliance on humans to complete this process.
Today, there are only 3-4 employees participating alongside this AI Agent, whereas the Customer previous had to rely on off-shore teams to supplement the manual effort. Although Roots has opted not to work within the management consulting and business services industry in favor of our preferred insurance and healthcare sectors, this remains a highly productive form of automation.