In this video we work through building an extractive Q&A stack using Haystack, and embedding it within a FastAPI instance in Python.

We use the BERT transformer for our reader model, alongside Elasticsearch and the BM25 retriever algorithm.