diff --git a/README.md b/README.md
index d0fce07..14a8a30 100644
--- a/README.md
+++ b/README.md
@@ -1,46 +1,28 @@
 # iFood Data Architect Test
 
-> This is my solution to the [iFood Data Architect Test](https://github.com/ifood/ifood-data-architect-test) 
+## Scope
+
+This is my solution to the [iFood Data Architect Test](https://github.com/ifood/ifood-data-architect-test) 
+
+> Work in progress! Solution not ready yet...
+
+## How to Run
+
+### Requirements
+
+* `docker >= 19.03.9`
+* `docker-compose >= 1.25.0`
+
+### Step-by-step
+
+* Run docker-compose at root directory
+
+```bash
+docker-compose up --build
+```
+
+* Run [notebooks](./spark-dev-env/docker-img-volume/notebooks) as you wish.
 
 ## Test Scope
 
-Process semi-structured data and build a datalake that provides efficient storage and performance. The datalake must be organized in the following 2 layers:
-* raw layer: Datasets must have the same schema as the source, but support fast structured data reading
-* trusted layer: datamarts as required by the analysis team
-
-![Datalake](./media/datalake.png)
-
-Use whatever language, storage and tools you feel comfortable to. 
-
-Also, briefly elaborate on your solution, datalake architecture, nomenclature, partitioning, data model and validation method.
-
-Once completed, you may submit your solution to ifoodbrain_hiring@ifood.com.br with the subject: iFood DArch Case Solution / Candidate Name.
-
-## Requirements
-
-* Source files:
-  * Order: s3://ifood-data-architect-test-source/order.json.gz
-  * Order Statuses: s3://ifood-data-architect-test-source/status.json.gz
-  * Restaurant: s3://ifood-data-architect-test-source/restaurant.csv.gz
-  * Consumer: s3://ifood-data-architect-test-source/consumer.csv.gz 
-* Raw Layer (same schema from the source):
-  * Order dataset.
-  * Order Statuses dataset.
-  * Restaurant dataset.
-  * Consumer dataset.
-* Trusted Layer:
-  * Order dataset -  one line per order with all data from order, consumer, restaurant and the LAST status from order statuses dataset. To help analysis, it would be a nice to have: data partitioned on the restaurant LOCAL date.
-  * Order Items dataset - easy to read dataset with one-to-many relationship with Order dataset. Must contain all data from _order_ items column.
-  * Order statuses - Dataset containing one line per order with the timestamp for each registered event: CONCLUDED, REGISTERED, CANCELLED, PLACED.
-* For the trusted layer, anonymize any sensitive data.
-* At the end of each ETL, use any appropriated methods to validate your data.
-* Read performance, watch out for small files and skewed data.
-
-## Non functional requirements
-* Data volume increases each day. All ETLs must be built to be scalable.
-* Use any data storage you feel comfortable to.
-* Document your solution.
-
-## Hints
-* Databricks Community: https://community.cloud.databricks.com
-* all-spark-notebook docker: https://hub.docker.com/r/jupyter/all-spark-notebook/
+Please find the code challenge [here](./TestScope.md).
\ No newline at end of file
diff --git a/TestScope.md b/TestScope.md
new file mode 100644
index 0000000..b7231a1
--- /dev/null
+++ b/TestScope.md
@@ -0,0 +1,44 @@
+# iFood Data Architect Test
+
+## Test Scope
+
+Process semi-structured data and build a datalake that provides efficient storage and performance. The datalake must be organized in the following 2 layers:
+* raw layer: Datasets must have the same schema as the source, but support fast structured data reading
+* trusted layer: datamarts as required by the analysis team
+
+![Datalake](./media/datalake.png)
+
+Use whatever language, storage and tools you feel comfortable to. 
+
+Also, briefly elaborate on your solution, datalake architecture, nomenclature, partitioning, data model and validation method.
+
+Once completed, you may submit your solution to ifoodbrain_hiring@ifood.com.br with the subject: iFood DArch Case Solution / Candidate Name.
+
+## Requirements
+
+* Source files:
+  * Order: s3://ifood-data-architect-test-source/order.json.gz
+  * Order Statuses: s3://ifood-data-architect-test-source/status.json.gz
+  * Restaurant: s3://ifood-data-architect-test-source/restaurant.csv.gz
+  * Consumer: s3://ifood-data-architect-test-source/consumer.csv.gz 
+* Raw Layer (same schema from the source):
+  * Order dataset.
+  * Order Statuses dataset.
+  * Restaurant dataset.
+  * Consumer dataset.
+* Trusted Layer:
+  * Order dataset -  one line per order with all data from order, consumer, restaurant and the LAST status from order statuses dataset. To help analysis, it would be a nice to have: data partitioned on the restaurant LOCAL date.
+  * Order Items dataset - easy to read dataset with one-to-many relationship with Order dataset. Must contain all data from _order_ items column.
+  * Order statuses - Dataset containing one line per order with the timestamp for each registered event: CONCLUDED, REGISTERED, CANCELLED, PLACED.
+* For the trusted layer, anonymize any sensitive data.
+* At the end of each ETL, use any appropriated methods to validate your data.
+* Read performance, watch out for small files and skewed data.
+
+## Non functional requirements
+* Data volume increases each day. All ETLs must be built to be scalable.
+* Use any data storage you feel comfortable to.
+* Document your solution.
+
+## Hints
+* Databricks Community: https://community.cloud.databricks.com
+* all-spark-notebook docker: https://hub.docker.com/r/jupyter/all-spark-notebook/
diff --git a/spark-dev-env/docker-img-volume/notebooks/02-ingest-raw-data.ipynb b/spark-dev-env/docker-img-volume/notebooks/02-ingest-raw-data.ipynb
new file mode 100644
index 0000000..1575d22
--- /dev/null
+++ b/spark-dev-env/docker-img-volume/notebooks/02-ingest-raw-data.ipynb
@@ -0,0 +1,52 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import boto3"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "cl = boto3.client('s3')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "cl.download_file()"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.8.5"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}