SKILL.md
506 lines · 1,802 words
Search Engineer
Purpose
Design and implement search systems that deliver fast, relevant results. This skill covers search engine selection, indexing strategies, relevance tuning, autocomplete, faceted search, and operational monitoring.
Search Engine Selection
Decision Matrix
ENGINE BEST FOR COMPLEXITY COST
------------------------------------------------------------
PostgreSQL FTS Simple search, < 1M docs Low Free
Meilisearch Typo-tolerant, instant Low Free/Hosted
Typesense Fast, easy setup Low Free/Hosted
Elasticsearch Complex queries, logs High Resource-heavy
OpenSearch AWS ecosystem High Resource-heavy
Algolia Hosted, instant UX Low Pay per search
DECISION TREE:
Do you need search beyond simple LIKE queries?
NO -> PostgreSQL LIKE with trigram index
YES -> How many documents?
< 100K -> Meilisearch or Typesense
100K - 10M -> Any engine (based on feature needs)
> 10M -> Elasticsearch / OpenSearch
Do you need:
Log aggregation + search? -> Elasticsearch
Instant autocomplete? -> Meilisearch or Typesense
Complex aggregations? -> Elasticsearch
Simple full-text? -> PostgreSQL FTS
Hosted with zero ops? -> Algolia