add redact

app.py

@@ -31,6 +31,8 @@ def process_msg(client, image):
             "image_url": f"data:image/jpeg;base64,{base_encode}" 
         }
     )
+    # in real model need to run Document AI first then redact after
+    # risk of AI seeing redacted info under black visual boxes
 
     processed_output=ocr_response.pages[0].markdown
 
@@ -50,8 +52,8 @@ def process_msg(client, image):
                     "and frequency (in medical short forms)," 
                     "dispense quantity,"
                     "and number of refills"
-                    "prescriber name, phone number and ID."
                     "return in json message only"
+                    "run this query twice and only return consistent information"
                     )
                 )
             ]

requirements.txt

@@ -1,3 +1,4 @@
 mistralai
 presidio-image-redactor
-pytesseract
+pytesseract
+https://github.com/explosion/spacy-models/releases/download/en_core_web_lg-3.7.1/en_core_web_lg-3.7.1-py3-none-any.whl