diff --git a/ingestion_server/env.template b/ingestion_server/env.template
index faf9be597a6..6cee2aafabd 100644
--- a/ingestion_server/env.template
+++ b/ingestion_server/env.template
@@ -4,8 +4,10 @@ PYTHONUNBUFFERED="0"
 
 #ENVIRONMENT="local"
 
-#AWS_ACCESS_KEY_ID=""
-#AWS_SECRET_ACCESS_KEY=""
+#AWS_REGION="us-east-1"
+#AWS_ACCESS_KEY_ID="test_key"
+#AWS_SECRET_ACCESS_KEY="test_secret"
+#AWS_S3_ENDPOINT="http://s3:5000"
 
 #ELASTICSEARCH_URL="es"
 #ELASTICSEARCH_PORT="9200"
diff --git a/ingestion_server/ingestion_server/cleanup.py b/ingestion_server/ingestion_server/cleanup.py
index 0fe8141de9a..d167eeb743d 100644
--- a/ingestion_server/ingestion_server/cleanup.py
+++ b/ingestion_server/ingestion_server/cleanup.py
@@ -13,8 +13,10 @@
 import uuid
 from urllib.parse import urlparse
 
+import boto3
 import requests as re
 import tldextract
+from decouple import config
 from psycopg2.extras import DictCursor, Json
 
 from ingestion_server.db_helpers import database_connect
@@ -302,6 +304,46 @@ def save_cleaned_data(result: dict) -> dict[str, int]:
     return cleanup_counts
 
 
+def _upload_to_s3(fields):
+    """
+    Upload cleaned data to S3. It assumes that the bucket already exists.
+
+    Locally, it connects to a MinIO instance through its endpoint and test credentials.
+    On live environments, the connection is allowed via IAM roles.
+    """
+    bucket_name = config("OPENVERSE_BUCKET", default="openverse-catalog")
+    s3_path = "shared/data-refresh-cleaned-data"
+    try:
+        s3 = boto3.resource(
+            "s3",
+            endpoint_url=config("AWS_S3_ENDPOINT", default=None),
+            aws_access_key_id=config("AWS_ACCESS_KEY_ID", default=None),
+            aws_secret_access_key=config("AWS_SECRET_ACCESS_KEY", default=None),
+            region_name=config("AWS_REGION", default=None),
+        )
+        s3.meta.client.head_bucket(Bucket=bucket_name)
+        bucket = s3.Bucket(bucket_name)
+        log.info(f"Connected to S3 and '{bucket_name}' bucket loaded.")
+    except Exception as e:
+        log.error(f"Files upload failed. Error connecting to S3.\n{e}")
+        return
+
+    for field in fields:
+        file_path = TMP_DIR.joinpath(f"{field}.tsv")
+        if not file_path.exists():
+            # Once the data has been cleaned in `upstream,` the cleaning process will
+            # not generate these files. Also, tags never generate any (refer to the
+            # `_clean_data_worker` function).
+            continue
+
+        try:
+            bucket.upload_file(file_path, f"{s3_path}/{field}.tsv")
+            log.info(f"Uploaded '{field}.tsv' to S3.")
+            file_path.unlink()
+        except Exception as e:
+            log.error(f"Error uploading '{field}.tsv' to S3: {e}")
+
+
 def clean_image_data(table):
     """
     Clean up data loaded from upstream that is unsuitable for prod before going live.
@@ -397,6 +439,7 @@ def clean_image_data(table):
     conn.commit()
     iter_cur.close()
     conn.close()
+    _upload_to_s3(cleanable_fields_for_table)
     end_time = time.perf_counter()
     cleanup_time = end_time - start_time
     log.info(