diff --git a/docs/deployments/realtime-api/predictors.md b/docs/deployments/realtime-api/predictors.md
index 97d89e702d..1869a5a133 100644
--- a/docs/deployments/realtime-api/predictors.md
+++ b/docs/deployments/realtime-api/predictors.md
@@ -532,3 +532,20 @@ def predict(self, payload):
         content=data, media_type="text/plain")
     return response
 ```
+
+## Chaining APIs
+
+It is possible to make requests from one API to another within a Cortex cluster. All running APIs are accessible from within the predictor at `http://api-<api_name>:8888/predict`, where `<api_name>` is the name of the API you are making a request to.
+
+For example, if there is an api named `text-generator` running in the cluster, you could make a request to it from a different API by using:
+
+```python
+import requests
+
+class PythonPredictor:
+    def predict(self, payload):
+        response = requests.post("http://api-text-generator:8888/predict", json={"text": "machine learning is"})
+        # ...
+```
+
+Note that the autoscaling configuration (i.e. `target_replica_concurrency`) for the API that is making the request should be modified with the understanding that requests will still be considered "in-flight" with the first API as the request is being fulfilled in the second API (during which it will also be considered "in-flight" with the second API). See more details in the [autoscaling docs](autoscaling.md).
diff --git a/pkg/workloads/cortex/serve/serve.py b/pkg/workloads/cortex/serve/serve.py
index c584156616..2d4265743d 100644
--- a/pkg/workloads/cortex/serve/serve.py
+++ b/pkg/workloads/cortex/serve/serve.py
@@ -21,6 +21,7 @@
 from concurrent.futures import ThreadPoolExecutor
 import threading
 import math
+import uuid
 import asyncio
 from typing import Any
 
@@ -121,7 +122,10 @@ async def register_request(request: Request, call_next):
     try:
         if is_prediction_request(request):
             if local_cache["provider"] != "local":
-                request_id = request.headers["x-request-id"]
+                if "x-request-id" in request.headers:
+                    request_id = request.headers["x-request-id"]
+                else:
+                    request_id = uuid.uuid1()
                 file_id = f"/mnt/requests/{request_id}"
                 open(file_id, "a").close()