diff --git a/pyproject.toml b/pyproject.toml
index 6fe9f63846..53e35e233f 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "feast-affirm"
-version = "0.28+affirm142"
+version = "0.28+affirm144"
 description = "Feast - Affirm"
 authors = ["Francisco Arceo", "Ross Briden", "Maks Stachowiak"]
 readme = "README.md"
diff --git a/sdk/python/feast/on_demand_feature_view.py b/sdk/python/feast/on_demand_feature_view.py
index 5b2c24b3f9..05aa23c299 100644
--- a/sdk/python/feast/on_demand_feature_view.py
+++ b/sdk/python/feast/on_demand_feature_view.py
@@ -362,9 +362,13 @@ def get_transformed_features_df(
                     dict_with_features[full_feature_ref] = dict_with_features[feature.name]
                     columns_to_cleanup.append(full_feature_ref)
         df_with_features = pd.DataFrame.from_dict(dict_with_features)
-        # Compute transformed values and apply to each result row
-        df_with_transformed_features = self.udf.__call__(df_with_features)
-
+        try:
+            # Compute transformed values and apply to each result row
+            df_with_transformed_features = self.udf.__call__(df_with_features)
+        except Exception as e:
+            with pd.option_context('display.max_rows', None, 'display.max_columns', None):
+                message = f"A udf error occured for {self.name}: {e}; udf inputs: {df_with_features}."
+                raise Exception(message) from e
         # Work out whether the correct columns names are used.
         rename_columns = self._get_correct_names(df_with_transformed_features.columns, full_feature_names)
 
@@ -391,15 +395,23 @@ def get_transformed_features_dict(
 
         # dict of lists into list of dicts
         input_rows = [dict(zip(feature_dict.keys(), values)) for values in zip(*feature_dict.values())]
-        output_rows = [self.udf.__call__(row) for row in input_rows]
+        output_rows = []
+        for row in input_rows:
+            for feature_value in row.values():
+                assert not isinstance(feature_value, pd.Series)
+            try:
+                output_rows.append(self.udf.__call__(row))
+            except Exception as e:
+                message = f"A udf error occured for {self.name}: {e}; udf inputs: {row}."
+                raise Exception(message) from e
         output_names = list(output_rows[0].keys())
         correct_names_map = self._get_correct_names(output_names, full_feature_names)
-        
+
         try:
             # list of dicts into dict of lists
             return {correct_names_map[name]: [row[name] for row in output_rows] for name in output_names}
         except KeyError as e:
-            logging.error(f"OnDemandFeatureView {self.name} udf returned unknown feature: {e.message}")
+            logging.error(f"OnDemandFeatureView {self.name} udf returned unknown feature: {e}")
             raise
 
     def infer_features(self):
diff --git a/setup.py b/setup.py
index de652b9975..11e9e0c032 100644
--- a/setup.py
+++ b/setup.py
@@ -39,7 +39,7 @@
     from distutils.core import setup
 
 NAME = "feast"
-VERSION = "0.28+affirm142"
+VERSION = "0.28+affirm144"
 DESCRIPTION = "Python SDK for Feast @ Affirm"
 URL = "https://github.com/feast-dev/feast"
 AUTHOR = "Feast"