Client Use Case: Automated Answering of PDF Questions
Background:
- A client has a use case involving over 300 questions in a PDF that need to be answered manually.
- Answers are based on a previously filled document.
- The PDFs come in different formats, and the questions may not exactly match.
Challenges:
- Identifying Answers:
- The first challenge was to identify the answers to these questions by referring to the previous document.
- Solution: AI technology was used to train a model that could provide answers based on the previous feed. This was successfully implemented.
- Updating the PDF with Answers:
- The second challenge was to update the PDF with answers once it is uploaded, given the varying formats from different clients.
- This challenge is not considered a good candidate for RPA (Robotic Process Automation) due to the format differences.