I have a mixed-mode survey that produced two very different kinds of answers: scanned handwritten notes for the qualitative questions and scanned images of check-box selections for the quantitative ones. My goal is to bring everything together in a single, tidy spreadsheet that I can immediately start analyzing. Here is what I need done: • Extract the handwritten comments using Azure Document Intelligence (or a comparable OCR approach in Python) and place each response in its own row, linked to the correct respondent ID. • Detect the ticked boxes in the scanned images, convert them to clear “0 / 1” or categorical values, and align them with the same respondent IDs. • Clean, de-duplicate, and format all fields so the final Excel file opens error-free and is ready for pivot tables or further statistical work. • Provide the Python scripts or notebooks you use so I can rerun the process on future batches. Acceptance is straightforward: if every respondent’s handwritten answer and checkbox data appear side-by-side in a single Excel workbook without missing or mis-aligned records, the job is complete. If turning messy inputs into clean, analysis-ready structure is your specialty, I look forward to seeing how you’d approach this.