Summary
In this chapter, we have discussed the fundamental tasks to build a neural search system, which are the indexing and querying pipelines. We looked into both of them and introduced the most challenging part, such as encoding and indexing.
You should have basic knowledge of the basic building blocks of indexing and querying, such as preprocessing, encoding, and indexing. You should also notice that the quality of the search results highly depends on the encoder, while the scalability of the neural search system highly depends on the indexer and the most popular algorithms behind the indexer.
As you need to build a production-ready search system, you will realize that purely relying on the basic building blocks is not enough. As a search system is complex to implement, it is always needed to design and add your own building blocks to the indexing and querying pipeline, in order to bring better search results.
In the next chapter, we will start to introduce Jina, the...
