Update README.md for split

README.md

@@ -84,6 +84,51 @@ we recommend using [python-dotenv](https://pypi.org/project/python-dotenv/)
 to add `VISION_AGENT_API_KEY="My Apikey"` to your `.env` file
 so that your Apikey is not stored in source control.
 
+### Split
+
+Split parsed documents into separate sections based on classification rules and identifiers.
+
+```
+import json
+import requests
+from io import BytesIO
+
+VA_API_KEY = 'YOUR_VA_API_KEY' # Replace with your API key
+headers = {"Authorization": f"Basic {VA_API_KEY}"}
+
+pdf_path = f'YOUR_PATH/TO/YOUR_PDF.pdf'
+
+# Parse the document first
+with open(pdf_path, "rb") as f:
+    parse_response = requests.post(
+        url="https://api.va.landing.ai/v1/ade/parse",
+        headers=headers,
+        files=[("document", f)],
+        # The selected model defaults to its latest available version.
+        # To specify an older version (e.g., 'model-name-date'), modify the model string.
+        # Details on all supported model versions: [https://docs.landing.ai/ade/ade-parse-models#model-versions-and-snapshots]
+        data={"model": "dpt-2"}
+    )
+
+# Split the document using the split rules
+split_class = [
+  {
+    "name": "Explanation of Benefits (EOB)",
+    "description": "A statement from a health insurance company detailing medical services received, amounts billed, and payments made.",
+    "identifier": "Claim #"
+  }
+]
+markdown_content = parse_response.json()["markdown"]
+split_response = requests.post(
+    url="https://api.va.landing.ai/v1/ade/split",
+    headers=headers,
+    files=[("markdown", BytesIO(markdown_content.encode('utf-8')))],
+    data={"split_class": json.dumps(split_class)},
+)
+
+print(split_response.json())
+```
+
 ### Parse Jobs
 
 For processing large documents asynchronously: