So much of our knowledge is recorded as textual data. The knowledge is there, but extracting insights out of it is a challenge. Imagine the time you spend trying to get to that one piece of information that you know is buried somewhere in your piles of documents. In this presentation, we will approach this problem by building our own semantic search application in Python, using Haystack. Haystack is an open source NLP framework and its key building blocks support a variety of semantic search pipelines. In this presentation, we will walk through one particular application of semantic search: question answering. We will also have a look at:

• What tasks semantic search enables
• Key building blocks
• How to leverage Haystack’s open source tooling to use the latest resources in NLP