PostgreSQL is (probably) all you need
summary
Claims that PostgreSQL can replace ES Redis Mongo and Neo4j get tested with specifics: full‑text using websearch_to_tsquery plus RUM and GIN indexes, ts_headline for snippets, a 300 GB subtitles and 20 GB lyrics corpus demo; graphs via recursive CTEs over 3 million nodes and 9 million links; JSONB document queries and PostgREST auto‑discovered joins; row level security policies for per‑row auth; extensions like pgvector and timescale and serverless hosting examples Neon and Supabase and practical tradeoffs for cache vs Redis.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 0:45Why PostgreSQL is underused
- 2:15PostgreSQL design philosophy
- 3:39Key-value store in PostgreSQL
- 4:45PostgreSQL vs Elasticsearch
- 7:13Full-text search indexing
- 10:50Query parsing for full-text search
- 13:47Demo: full-text search
- 16:43Trees and graphs with CTEs
- 21:00Graph traversal and social networks
- 25:21JSON documents in PostgreSQL
- 31:27Backend API with PostgREST
- 33:28Row-level security
- 36:22PostgreSQL extensions
- 37:50Hosting and serverless options
- 38:33Conclusion and Q&A
related talks
43mWhy Postgres
29mPostgres Demystified
25mThe Rise of PostgreSQL as the Everything Database
50mScaling Postgres With Some Help from Redis
42mPostgreSQL Tips & Tricks For App Devs
51mHow Postgres is Different (read BETTER than) your database
43mJust Use Postgres.
33mPostgres, the Best Tool You're Already Using