Get comfortable with full-text and vector search in MongoDB Atlas.
0 / 5 completed
1 / 5
At standup, a dev enables full-text search on a collection in Atlas. What must they create first?
Atlas Search requires a search index built on the collection, backed by Lucene. The index defines analyzers and field mappings. Without it, the $search stage cannot run.
2 / 5
During a design review, a dev adds search to an aggregation pipeline. Which stage do they use?
The $search stage must be the first stage of an aggregation pipeline and runs queries against an Atlas Search index. Subsequent stages can refine or shape results. It is the core entry point for Atlas Search queries.
3 / 5
In a code review, a dev combines multiple conditions with must/should/filter. Which operator is this?
The compound operator combines clauses like must, should, mustNot, and filter to build complex relevance logic. It mirrors boolean query composition. This is how you express multi-condition searches.
4 / 5
An incident report needs counts per category alongside results. Which Atlas Search feature provides this?
Atlas Search facets return aggregated counts per bucket (e.g., per category) alongside search hits. They power filter sidebars in search UIs. You declare facet definitions in the search stage or index.
5 / 5
During a PR review, a dev adds vector similarity search. Which Atlas Search operator was used historically?
Vector similarity in Atlas Search was introduced via the knnBeta operator for k-nearest-neighbor queries over embedding fields. It enables semantic search alongside text. Newer vector search interfaces have since built on this foundation.