Life Sciences, Healthcare, Vials, Commercialization
Case Study

Manufacturer identifies discrepancies in employee benefit plans payments

Learn how a leading manufacturer utilized record linkage algorithm and data science to effectively identify and rectify late contributions to retirement plans.
Life Sciences, Healthcare, Vials, Commercialization
Case Study

Manufacturer identifies discrepancies in employee benefit plans payments

Learn how a leading manufacturer utilized record linkage algorithm and data science to effectively identify and rectify late contributions to retirement plans.

Client background

The company is a multinational manufacturer of drug packaging products with hundreds of employees in the United States.

The business challenge

The manufacturer needed assistance in identifying late contribution remittances to their employees’ benefit plans and confirming the length of the untimely remittances, ensuring that necessary corrections were made.

However, the records from the payroll provider and the third-party custodian did not entirely match. There was no common identifier between the two, and the dates of paycheck withholding and fund receipt did not coincide. This discrepancy made it difficult to match the two datasets to identify the length of the untimely remittances.

Strategy and solution

Baker Tilly’s digital team, in collaboration with Baker Tilly’s employee benefit plan practice took the following approach to identify the length of the untimely remittances:

  1. Conduct an assessment and formulate analytic questions: A thorough data examination and preparation phase tailored to meet the company’s objectives allowed us to understand the problem in depth and formulate precise, practical analytic questions for testing.
  2. Collect data: Our data collection process involved identifying significant patterns in paycheck withholding and fund receipt data as well as identifying employees with identical names in order to gather additional information needed to distinguish them. We then transformed columns of diverse formats into a unified format to streamline the data. Recognizing the sensitive nature of the data, we conducted our analysis using AWS SageMaker and AWS Athena, prioritizing security and efficiency.
  3. Design and execute testing procedures: Based on the similarity of employee names, dates of payroll and remittance transactions and transaction amounts, we designed and implemented a machine learning algorithm to perform record linkages. To ensure the integrity of our process, we conducted thorough quality checks to verify the accuracy of the matches.
  4. Record testing procedures and results: We maintained transparency by providing a clear and precise depiction of our results. We also scrutinized the client’s feedback on the matching results and rigorously tested it to ensure the majority of the unmatched records were adequately addressed.
  5. Provide a narrative report on findings: We delivered a narrative report that encompassed intuitive explanations of the results. These clear and concise explanations enable the company’s management to make informed decisions and implement necessary corrections based on our findings.

As a result of this engagement, the company was able to:

  • Reduce the manual effort required to examine each employee record for identifying untimely remittances, thereby saving valuable time and resources
  • Enhance the efficiency and accuracy of identifying late contributions, ensuring that no discrepancies were overlooked
  • Enable faster and accurate remedial actions to be taken, minimizing the impact of late contributions
  • Ensure secure data handling methods, establishing the highest level of protection for confidential information

Chatting with artificial intelligence chatbot on mobile device
Next up

Software solution provider enhances customer service and product knowledge with generative AI