diff --git a/scripts/restore-postgres/README.md b/scripts/restore-postgres/README.md
new file mode 100644
index 000000000..932eef42d
--- /dev/null
+++ b/scripts/restore-postgres/README.md
@@ -0,0 +1,23 @@
+# Restore backup from S3
+
+## Copy file from S3 to local env
+
+```
+aws s3 cp s3://<bucketname>/<exported-data-directory> </local/path> --recursive --profile <AWS_PROFILE>
+```
+
+## Install dependencies
+
+```
+pip install pandas pyarrow psycopg2 sqlalchemy
+```
+
+## Create database through MartenDb
+
+Eg: start the services once on your local dev machine, stop them after db creation.
+
+## Run restore.py
+
+```
+python restore.py
+```
diff --git a/scripts/restore-postgres/restore.py b/scripts/restore-postgres/restore.py
new file mode 100644
index 000000000..008f8367f
--- /dev/null
+++ b/scripts/restore-postgres/restore.py
@@ -0,0 +1,46 @@
+import os
+import pandas as pd
+from sqlalchemy import create_engine
+import logging
+
+# Database connection settings
+db_username = 'username'
+db_password = 'password'
+db_host = '127.0.0.1'
+db_port = '5432'
+db_name = 'verenigingsregister'
+
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+
+
+# PostgreSQL engine
+engine = create_engine(f'postgresql://{db_username}:{db_password}@{db_host}:{db_port}/{db_name}')
+
+# Directory containing Parquet files
+parquet_dir = '/path/to/verenigingsregister'
+
+def format_table_name(path):
+    parts = path.split(os.sep)
+    if len(parts) > 1:
+        # Extracts the part after 'public.'
+        schema_table = parts[-2]
+        if schema_table.startswith('public.'):
+            return schema_table.split('.')[1]
+    return None
+
+# Function to load a Parquet file into PostgreSQL
+def load_parquet(file_path, table_name):
+    df = pd.read_parquet(file_path)
+    df.to_sql(table_name, engine, if_exists='append', index=False)
+
+# Loop through directories and files
+for root, dirs, files in os.walk(parquet_dir):
+    for file in files:
+        if file.endswith('.gz.parquet'):
+            file_path = os.path.join(root, file)
+            # Extract and format table name
+            table_name = format_table_name(root)
+            if table_name:
+                load_parquet(file_path, table_name)
+
+print("Data import complete.")
\ No newline at end of file