Senior UiPath Developer
H1 Life Sciences Pvt Ltd · Sep 2021 – Apr 2025
- Project: HCA healthcare -invoice validation) we were processing large volumes of unstructured healthcare documents like claims and invoices from multiple hospital systems.
- The manual process was time-consuming, error-prone, and lacked scalability.
- I worked on a healthcare Tungsten automation project where we had to process large volumes of unstructured documents like claims and invoices coming from different hospital systems.
- The main challenge was that these documents had no fixed format, so traditional automation was not effective.
- To solve this, I designed an Intelligent Document Processing solution using UiPath Document Understanding.
- I used OCR engines like ABBYY to digitize the documents, and then implemented ML-based classification and extraction using UiPath AI Center.
- Since the documents were highly unstructured, I followed a hybrid approach-combining ML Extractor, Regex Extractor, and Form Extractor to improve accuracy.
- I also implemented human-in-the-loop validation using Validation Station and Action Center, so that low-confidence fields could be reviewed by users and the corrected data could be used for retraining the model.
- Additionally, I integrated Python for advanced data processing tasks like cleaning OCR output, restructuring table data, and performing fuzzy matching for validation.
- I also explored Generative AI using Python APIs for document summarization and handling edge cases where traditional extraction was not sufficient.
- The entire solution was built on REFramework with queue-based processing, which allowed us to scale and handle high volumes efficiently.
- As a result, we achieved around 85 to 90 percent automation, reduced manual effort by more than 60 percent, and greatly improved data accuracy over time.
- Overall, it was a strong combination of RPA, AI, and human validation to handle complex unstructured data.

