diff --git a/validator/main.py b/validator/main.py
index 759d0b4..0777e3c 100644
--- a/validator/main.py
+++ b/validator/main.py
@@ -37,7 +37,7 @@ def __init__(
         self._arg_1 = arg_1
         self._arg_2 = arg_2
 
-    def validate(self, value: Any, metadata: Dict = {}) -> ValidationResult:
+    def _validate(self, value: Any, metadata: Dict = {}) -> ValidationResult:
         """Validates that {fill in how you validator interacts with the passed value}."""
         # Add your custom validator logic here and return a PassResult or FailResult accordingly.
         if value != "pass": # FIXME
@@ -46,3 +46,42 @@ def validate(self, value: Any, metadata: Dict = {}) -> ValidationResult:
                 fix_value="{The programmtic fix if applicable, otherwise remove this kwarg.}",
             )
         return PassResult()
+
+    def _inference_local(self, model_input: Any) -> Any:
+        """
+        Runs a machine learning pipeline on some input on the local
+        machine. This function should receive the expected input to the
+        ML model, and output the results from the ml model.
+        """
+        raise NotImplementedError
+
+    def _inference_remote(self, model_input: Any) -> Any:
+        """
+        Runs a machine learning pipeline on some input on a remote
+        machine. This function should receive the expected input to the
+        ML model, and output the results from the ml model.
+
+        """
+        raise NotImplementedError
+
+    def _inference(self, model_input: Any) -> Any:
+        """Calls either a local or remote inference engine for use in the
+        validation call.
+
+        Args:
+            model_input (Any): Receives the input to be passed to your ML model.
+
+        Returns:
+            Any: Returns the output from the ML model inference.
+        """
+        # Only use if both are set, otherwise fall back to local inference
+        if self.use_local:
+            return self._inference_local(model_input)
+        if not self.use_local and self.validation_endpoint:
+            return self._inference_remote(model_input)
+
+        raise RuntimeError(
+            "No inference endpoint set, but use_local was false. "
+            "Please set either use_local=True or "
+            "set an validation_endpoint to perform inference in the validator."
+        )