diff --git a/scripts/airflow_job.py b/scripts/airflow_job.py
index b1b6c94..4260db7 100644
--- a/scripts/airflow_job.py
+++ b/scripts/airflow_job.py
@@ -10,7 +10,7 @@ from datetime import date
 from datetime import datetime
 from datetime import timedelta
 
-from taar_loader import runme
+import taar_loader
 
 
 def parse_args():
@@ -41,4 +41,4 @@ def parse_args():
 
 if __name__ == "__main__":
     args = parse_args()
-    runme(args.run_date)
+    taar_loader.runme(args.run_date)
diff --git a/taar_loader/__init__.py b/taar_loader/__init__.py
index 33d6ee3..36a348b 100644
--- a/taar_loader/__init__.py
+++ b/taar_loader/__init__.py
@@ -1 +1,2 @@
 from taar_dynamo import runme
+from filters import dynamo_reducer
diff --git a/taar_loader/filters.py b/taar_loader/filters.py
index 67e66a8..120c662 100644
--- a/taar_loader/filters.py
+++ b/taar_loader/filters.py
@@ -1,6 +1,7 @@
 # This Source Code Form is subject to the terms of the Mozilla Public
 # License, v. 2.0. If a copy of the MPL was not distributed with this
 # file, You can obtain one at http://mozilla.org/MPL/2.0/.
+import boto3
 
 
 def filterDateAndClientID(row_jstr):
@@ -10,7 +11,7 @@ def filterDateAndClientID(row_jstr):
         assert row.client_id is not None
         dateutil.parser.parse(row.subsession_start_date)
         return True
-    except Exception as inst:
+    except Exception:
         return False
 
 
@@ -35,33 +36,40 @@ def list_transformer(row_jsonstr):
     return (0, 1, [jdata])
 
 
-def boto3_tuple_reducer(tuple_a, tuple_b):
-
-    if tuple_a[1] == 0:
-        return tuple_b
-    if tuple_b[1] == 0:
-        return tuple_a
-
-    # Both tuples have non-zero length lists, merge them
-
-    working_tuple = [tuple_a[0] + tuple_b[0],
-                     tuple_a[1] + tuple_b[1],
-                     tuple_a[2] + tuple_b[2]]
-
-    if working_tuple[1] >= 3:
-        push_to_dynamo(working_tuple[2])
-        working_tuple[0] += working_tuple[1]
-        working_tuple[1] = 0
-        working_tuple[2] = []
-
-    return tuple(working_tuple)
-
-
-def push_to_dynamo(json_list):
-    import boto3
+def push_to_dynamo(item_list):
+    """
+    This connects to DynamoDB and pushes records in `item_list` into
+    a table.
+    """
     conn = boto3.resource('dynamodb', region_name='us-west-2')
     table = conn.Table('taar_addon_data')
-
     with table.batch_writer(overwrite_by_pkeys=['client_id']) as batch:
-        for item in json_list:
+        for item in item_list:
             batch.put_item(Item=item)
+
+
+def dynamo_reducer(list_a, list_b):
+
+    # If one list is empty, just return the other list
+    if list_a[1] == 0:
+        return list_b
+    if list_b[1] == 0:
+        return list_a
+
+    new_list = [list_a[0] + list_b[0],
+                list_a[1] + list_b[1],
+                list_a[2] + list_b[2]]
+
+    if new_list[1] > 100:
+        push_to_dynamo(new_list[2])
+
+        # Update number of records written to dynamo
+        new_list[0] += new_list[1]
+
+        # Zero out the number of accumulated records
+        new_list[1] = 0
+
+        # Clear out the accumulated JSON records
+        new_list[2] = []
+
+    return tuple(new_list)
diff --git a/taar_loader/taar_dynamo.py b/taar_loader/taar_dynamo.py
index 8e0cc1b..b7e63c7 100644
--- a/taar_loader/taar_dynamo.py
+++ b/taar_loader/taar_dynamo.py
@@ -13,7 +13,8 @@ from pyspark.sql import SparkSession
 from pyspark.sql.functions import desc, row_number
 from pyspark.sql import Window
 
-from taar_loader.filters import filterDateAndClientID, list_transformer
+from taar_loader.filters import filterDateAndClientID
+from taar_loader.filters import list_transformer
 
 
 def etl(spark, run_date, dataFrameFunc):
@@ -91,7 +92,8 @@ def main(spark, run_date):
     def load_parquet(dateString):
         return spark.read.parquet(template % dateString)
 
-    return etl(spark, run_date, load_parquet)
+    rdd = etl(spark, run_date, load_parquet)
+    return rdd
 
 
 def reducer(tuple_a, tuple_b):