Helpline router


  • S3 keeps full text search index of topic (~10-20)
  • Lambda process the user input using the index
  • Lex to get user input


  • user call the helpline, and is asked to say what he wants to talk about
  • based on user input (in text), a search on full text index is one to determine the topic
  • when the topic is know user is routed to a proper consultant

Extra (if there is time)

  • Machine learning can be applied to better determine if user input matches the topic (classification based on user input)

Built With

Share this project: