In June 2020, Twitter announced a major overhaul in its storage and retrieval systems. These changes allowed Twitter to reduce the search index latency from 15 seconds to 1 second.

So, what did they do to get such impressive gains? Allow me to explain!

Word vector representations with subword information are great for NLP modeling. But can we make lexical corrections using a trained embeddings space? Can its accuracy be high enough to beat Peter Norvig’s spell-corrector? Let’s find out!