diff --git a/README.md b/README.md
index 531eb36428eaf541c2d06621adfaad71dae5f413..efe1da214140e83354251c45ed20f4501ea72b04 100644
--- a/README.md
+++ b/README.md
@@ -1,5 +1,16 @@
-# UnicoreSpawner
 
-Spawner enables start of singleuser notebook-servers via [UNICORE](https://www.unicore.eu) on various HPC-Systems.  
+[![Documentation Status](https://readthedocs.org/projects/jupyterhub-unicorespawner/badge/?version=latest)](https://jupyterhub-unicorespawner.readthedocs.io/en/latest/?badge=latest)
+
+# UNICORESpawner
+
+The UNICORESpawner enables JupyterHub to spawn services via [UNICORE](https://www.unicore.eu).
+It is a subclass of the [JupyterHub ForwardBaseSpawner](https://github.com/kreuzert/jupyterhub-forwardbasespawner).
+
+## Features
+
+UNICORESpawner combines the power of UNICORE with the simplicity of JupyterHub. Using the UNICORE REST-API, the UNICORESpawner can spawn services on any remote system connected to UNICORE/X. 
+  
+- Start jupyter notebook servers on hpc systems.
+- Communication with jupyter server in a browser, even if they're running on batch nodes without internet access, enables "Supercomputing in a browser".
+- Use Callback feature of UNICORE reduces poll overload. JupyterHub will be informed, if the stauts of the UNICORE job changes.
   
-Used at Juelich supercomputing centre for https://jupyter-jsc.fz-juelich.de and other JupyterHub installations.