This document discusses how LinkedIn's internal learning portal LearnIn was developed using Node.js, MongoDB, and Elasticsearch. It covers using Mongoose to model schemas and build APIs in MongoDB. To enable full-text search, an Elasticsearch river was configured to replicate data from MongoDB to Elasticsearch for improved search capabilities like relevancy scoring. The document provides code examples of setting up Mongoose schemas, queries, and configuring the Elasticsearch river.