Microsofts' LUIS service will easily do that for you. You train it to recognize a user intent (show_invoices, record_payment) and classify additional entities (payment_status, creditor_name, or payment_type). It works remarkably well.
You use the portal to register your application and enter some example sentences and specify your expected interpretation. Then you can start recognizing input strings using the API. You can then improve the recognizer by manually correcting input from actual use.
Have a look at 'Named Entity Recognition' (NER), particularly a 7-class model to recognize: location, person, organization, money, percent, date and time.
The Stanford NLP Group has a NER project: http://www-nlp.stanford.edu/software/CRF-NER.shtml, and there are many others.
Are there any libraries or platforms that would help me implement this kind of natural language UI?