Worked on the mindsandcompany/doc_parser repository to enhance the doc_parser pipeline by introducing a configurable EasyOCR recognition network option. This feature allows users to select the recognition model through a new recog_network parameter, providing greater flexibility and control over OCR accuracy and tuning. The implementation involved updates to pipeline_options.py and easyocr_model.py, ensuring seamless integration with existing backend logic. Leveraging Python for backend development and API integration, the work focused on expanding customization options without disrupting default behaviors. Over the course of the month, the contribution centered on feature development, with no bug fixes, emphasizing targeted improvements to OCR configurability.