izhangzhihao implemented two custom components that we wanted to share with the community. You can find the implementation here: GitHub - izhangzhihao/rasa-contrib: Third party conponents of rasa
The first component tokenizes a text by making a request to an external server. Thus, you can, for example, set up a Standford CoreNLP server and use the component of izhangzhihao to get incoming text tokenized by Standford CoreNLP. The other component does the same for entity extraction. You can extract entities from a text using a third party extractor that is called via a HTTP request.