5 mins for Every Request (~50 Requests per week)
~ 2 mins every Request
Challenge: Our client’s human resources department faced the challenge of validating education claims requests submitted by employees. These requests required verification of PDF receipts provided by education institutions. Only authentic and valid receipts could be approved for reimbursement, making the validation process crucial. However, the manual review of PDF receipts was time-consuming and prone to errors.
Solution: To address the challenge of validating education claims, we implemented Automation Anywhere 360 to automate the entire process. Our solution aimed to streamline the verification process, reduce manual errors, and enhance the overall user experience.
Process: The RPA bot executed the following steps to automate the verification of education claims PDFs:
- Oracle Fusion ERP Login: The bot logged into the Oracle Fusion ERP system, gaining access to the necessary modules and functionalities.
- Pending Requests: The bot retrieved the list of pending education claims requests from the ERP system. It created a transactions file containing the details of these requests and loaded the information into the database.
- Receipt Download: The bot downloaded the PDF receipts associated with each request from the ERP system. These receipts were obtained from education institutions and submitted by employees for reimbursement.
- Data Analysis using Chat GPT: The bot utilized Chat GPT, an AI model, to analyze the data extracted from the uploaded receipts. It extracted key details such as the name of the educational institution, the program/course name, and the duration of the program. This analysis helped in verifying the authenticity of the claims.
- Claim Approval or Rejection: Based on predefined conditions stored in a master configuration file, the bot made decisions regarding claim approval or rejection. If the receipts were deemed valid and met the predefined criteria, the bot approved the claim. Otherwise, it rejected the claim and provided a reason for rejection.
- Process Completion: The bot repeated the above steps for all pending education claims requests. It saved the final transaction file containing the updated status of each request, along with transaction logs and screenshots, in a shared folder. Additionally, the bot sent a notification to the relevant stakeholders, providing the final transaction file, transaction logs, and any business exceptions that required attention.
- Daily Execution: The bot performed this automated task daily, ensuring that the verification of education claims PDFs was conducted promptly and efficiently.