For now, end users can achieve 2 things: 1. Context aware, semantically correct answers to questions (based on the user's content). This is ideal for automatic virtual assistant, FAQ answering via a bot etc 2. Context aware, semantically correct conversations (chat like) for expert systems (i.e. Natural language interface for an expert system).

I'm really excited about the accuracy I'm seeing and want to release it to some real world clients.