Transforming Loan Application Processing with Butler's AI-Driven Document Extraction: A Lendtable Case Study

Posted by:
Matt Noe
on
April 25, 2023

Introduction

Lendtable, a fintech startup, is revolutionizing the way employees access their company benefits by offering cash advances based on their benefits package. This innovative approach empowers individuals to maximize their financial potential by leveraging their employer's 401K plan or ESPP. 

To streamline the application process, applicants only need to upload their paystub. Information from these documents is then used to determine whether an advance should be approved or not, a task that demands efficiency and accuracy. 

In this case study, we will explore how Butler's cutting-edge document extraction solution has transformed Lendtable's workflow, resulting in significant time savings and improved customer experience.

Challenge

Lendtable faced several challenges when processing applications that relied on extracting information from unstructured documents, such as paystubs and 401K statements. 

Their existing process relied on humans and was slow. It would take up to 24 hours for customers to be notified of a term sheet to review. 

This delay negatively impacted the customer experience and led to a loss of revenue as some potential clients abandoned the process. 

Lendtable’s existing document extraction provider was not a good candidate for reviewing paystub applications due to the following reasons:

  • Had slow response times (usually minutes to hours) when Lendtable was seeking a solution that would review a paystub and generate a term sheet to a customer in mere seconds. 
  • Did not offer the ability to fine-tune the paystub extraction models for their specific use case, resulting in recurring errors that required manual review and data entry.
  • The costs associated with their existing provider were prohibitive, putting additional strain on the business.

Solution

To overcome the challenges faced in their customer application process, Lendtable turned to Butler's managed large language models to create a tailored solution. They worked with the Butler team to fine-tune large language models specifically for paystub extraction, enabling them to address the unique requirements of their use case. 

These models offered several key benefits:

  • Real-time extraction: By leveraging Butler's advanced AI technology, Lendtable was able to process unstructured documents in real-time, significantly reducing the time required for account approvals.
  • High accuracy: The custom extraction models boasted an accuracy rate of over 95%, ensuring that Lendtable could rely on the extracted data with minimal manual intervention.
  • Easy integration: Butler's easy-to-use REST API made it simple for Lendtable to integrate the document extraction solution into their existing workflow, allowing for a seamless transition.

Yuri, a lead product manager on the Lendtable team, shared about their experience with Butler: 

"Butler's fine-tuned large language models have been a game changer for our customer application process. The real-time extraction, high accuracy, and seamless integration into our existing workflow have led to faster application times, increased revenue, and reduced manual review. We couldn't be happier with the results."

Results

After implementing Butler's custom document extraction solution, Lendtable experienced impressive results that positively impacted their business:

  • Faster application times: Thanks to real-time extraction, the time required for customer applications was reduced by 5x, cutting down the process from hours to mere minutes.
  • Increased revenue: The accelerated application turnaround allowed Lendtable to retain more potential customers, leading to a boost in revenue as fewer clients dropped off during the process.
  • Operational efficiencies from less manual review: The fine-tuned models tailored for Lendtable's use case resulted in fewer exception documents, reducing the need for manual review and increasing overall operational efficiency.
  • Minimal engineering effort: The seamless integration of Butler's easy-to-use APIs made it possible for Lendtable to achieve these significant improvements with minimal engineering effort.

The combination of these results greatly enhanced Lendtable's onboarding application process, customer experience, and overall business performance.

Conclusion

In conclusion, Butler's cutting-edge document extraction solution played a pivotal role in transforming Lendtable's customer application process. By fine-tuning large language models for their specific document extraction use case, Lendtable significantly reduced application times, increased revenue, and improved operational efficiency with minimal engineering effort. The successful partnership between Lendtable and Butler showcases the power of AI-driven solutions in revolutionizing the fintech industry and enhancing customer experiences.

As Lendtable continues to innovate and expand its offerings, the company looks forward to exploring new opportunities with Butler, including the use of their advanced AI technology for 401K and ESPP plan analysis and other potential applications. This ongoing collaboration is a testament to the value that Butler's solutions bring to the rapidly evolving fintech landscape.

Build document extraction into your product or workflow today!