Scanning Service Bureau
Case Study
Learn more about how JetStream can work in real situation with this scanning service bureau case study. In which the client, the client's challenge and the potential solution in our JetStream software becomes clear
The Client
High volume client
This respected client has been providing business process outsourcing services for over five decades. They have gained a reputation for their expertise in serving healthcare providers, the public sector, and enterprise customers.
With a strong focus on digital transformation, they specialize in digitizing, indexing, and archiving various documents. Impressive in scale, they process a staggering volume of one billion pages each year.
240-Day
Set-up Time of Previous Solution
50%
Current Position Automation Rate
The Challenge
Low and Inefficient Automation Rates
Prior to implementing JetStream, the client had a daily process that involved scanning and reading incoming boxes of documents. They initially used a popular Intelligent Document Processing (IDP) product for rule-based document classification, however, they faced significant challenges in automating their scanning and archiving processes. Their automation rate for document classification was only 50%, which meant a large portion of manual work was needed to correct and validate documents that didn't meet the necessary confidence thresholds.
This customer’s rule-based document classification approach was resource-intensive and costly to maintain. Making changes to existing document classes or creating new ones was particularly burdensome. The introduction of regulatory changes highlighted the need for a more precise categorization system that their previous solution couldn't effectively provide.
The Service Bureau struggled with low automation rates, manual validation work, high maintenance costs, and the inability to handle regulatory changes with their existing document classification process. They sought a more efficient and accurate solution, which led them to consider implementing JetStream.
The Solution
Using JetStream to increase efficiency and automation
Automation for Document Classification
During the evaluation phase, the implementation of JetStream demonstrated exceptional performance. Right out of the box, it achieved an impressive automation rate of 90% for document classification. This accomplishment can be credited to JetStream's cutting-edge OCR engine, which provides remarkable accuracy and confidence scores.
Machine Learning & Quick Setup
JetStream's rule-free, few-shot learning approach, sets it apart from the previous solution. This enabled a swift initial setup, taking less than 3-days, a significant improvement compared to the lengthy 240-day setup time required by the previous solution. The client faced regulatory changes that necessitated a substantial increase in document classes, going from approximately 150 to over 300. JetStream's ability to quickly adapt to these new document classes allowed the client to quickly meet their requirements.
Reduction in Manual Effort & Minimizing False Positives
The implementation of JetStream had a substantial impact, resulting in an 80% reduction in manual efforts. This was achieved through increased straight-through processing, even with the client's objective of minimizing false positive results to less than 1%. The significant improvement in automation provided by JetStream empowers our client to scale their business significantly.
The Benefits & Results of JetStream Implementation
Outlook
Growth Through JetStream Automation
Processing large batches of documents, especially PDFs containing 100 or more consecutive forms, can be extremely challenging. Automating the classification of these documents based on their categories may seem nearly impossible. Currently, the process of separating these documents requires a significant amount of manual effort, involving the insertion of separator sheets between each document.
However, looking to the future, the client plans to tackle this issue by utilizing JetStream's document splitting feature. This feature will allow them to train a neural network to automatically identify the start and end of multi-page documents or split documents based on a fixed number of pages using rule-based methods.
By taking advantage of JetStream's data extraction capabilities, the client will have the opportunity to expand their service portfolio and explore further growth opportunities.
