diff --git a/README.md b/README.md
index 38bbc16..8a7044f 100644
--- a/README.md
+++ b/README.md
@@ -4,30 +4,50 @@ Run milobella in a device, using mic and speakers to communicate.
 
 > Disclaimer : Only checked in a Raspberry PI 3B+ for now.
 
-## Speech-To-Text
+# Technical choices
+#### Speech-To-Text
 The translation of voice records into text is currently performed by the [Google Cloud Speech-To-Text API](https://cloud.google.com/speech-to-text).
+
 The [mozilla/deepspeech](https://github.com/mozilla/DeepSpeech) project with Common Voice database is considered.
 
-## Text-To-Speech
-The translation of Milobella text answers into speech synthesis is ensured by [Google Cloud Text-To-Speech API](https://cloud.google.com/text-to-speech)
+#### Text-To-Speech
+The translation of Milobella text answers into speech synthesis is ensured by [Google Cloud Text-To-Speech API](https://cloud.google.com/text-to-speech).
+
 No replacement has been considered for now.
 
-## Google cloud authentication configuration
-Authentication process is detailed in this documentation [https://cloud.google.com/speech-to-text/docs/quickstart-protocol].
+# Requirements
+#### Google cloud authentication configuration
+Authentication process is detailed in this documentation https://cloud.google.com/speech-to-text/docs/quickstart-protocol.
+
 It is using basically a ``GOOGLE_APPLICATION_CREDENTIALS`` env variable pointing to a JSON private key file.
 The JSON private key is generated when you create a Service Account.
 
 Make sure you activated both of the APIs in your Google Cloud Project, and the same JSON file will be used.
 
-## Milobella authentication configuration
+#### Milobella authentication configuration
 The env variable ``MILOBELLA_AUTHORIZATION_TOKEN`` should contain the JWToken generated every time you authenticate.
 A script ``authenticate.sh`` is here to show you how. (You need [jq](https://stedolan.github.io/jq/download/) if you want to use it)
 ```
+export MILOBELLA_USERNAME=myuser
+export MILOBELLA_PASSWORD=mypass
 source authenticate.sh
 ```
 
-## Raspberry audio configuration
-If you are using a Raspberry PI, an example of ``.asoundrc`` is located in this repository. We use plugin to
-transform the sample rate to 16GHz. Somehow it works with this configuration.
+#### Raspberry audio configuration
+If you are using a Raspberry PI, an example of ``.asoundrc`` is located in this repository. It is using a plugin to
+transform the default sample rate to 16GHz, which is necessary to make work the google cloud speech-to-text. Somehow it works with this configuration ...
+
+The file ``.asoundrc`` needs to be copied in $HOME directory to be effective, and probably some adjustments are necessary to match the proper card and devices.
+
+# Run the program
+```
+# [If not already in a venv] Not mandatory but it is always easier to have a virtualenv
+python3 -m venv venv
+source venv/bin/activate
+
+# Install requirements
+pip install -r requirements.txt
 
-The file ``.asoundrc`` needs to be copied in $HOME directory to be effective.
+# Run the app
+python milobella.py
+```