Blog
PostgreSQL, pgvector, and AI infrastructure: guides and deep dives.
Filter by tag
All posts
PostgreSQL Full Text Search: A Complete Guide
Learn how PostgreSQL full text search works, from tsvector and tsquery basics to GIN indexes and ranking. Covers real SQL, performance tips, and FTS vs pgvector.
9 min readPostgreSQL High Availability: A Practical Guide for Production
Learn how PostgreSQL high availability works, compare Patroni vs pg_auto_failover, and decide when to manage HA yourself vs use a managed service.
10 min readHow to Use pgvector with Python: A Complete Guide
Learn how to use pgvector with Python to store and query vector embeddings in PostgreSQL. Covers setup, psycopg3, SQLAlchemy, and building a real RAG pipeline.
8 min readpgvector vs Pinecone: Which Should You Use in 2026?
pgvector vs Pinecone compared on performance, cost, and developer experience. Find out which vector database fits your AI stack in 2026.
10 min read7 Supabase Alternatives Compared (2026): Pricing and pgvector
Seven managed Postgres platforms compared on real monthly cost, pgvector support, and who each is actually for. Including when to just stay on Supabase.
14 min readpgvector Benchmarks, Per Tier: Honest, Reproducible Numbers from Starter to Scale
Exactly how our dedicated nodes run pgvector HNSW search, measured honestly per tier (Starter 2vCPU/4GB, Growth 4vCPU/8GB, Scale 8vCPU/16GB) on 250k to 1M × 1536 vectors, with the QPS/recall/latency tradeoffs, the build-memory ceiling, and a CLI to reproduce every number yourself.
17 min readPostgreSQL 18: What's New and Why It Matters
PostgreSQL 18 brings async I/O, faster JSONB, smarter partitioning, and better replication. Here's what actually matters for production workloads and AI applications.
6 min readGetting Started with pgvector on Rivestack
A practical guide to setting up pgvector for AI-powered similarity search on managed PostgreSQL. From first query to production-ready indexes in one sitting.
7 min readWhy Managed PostgreSQL is the Best Choice for AI Workloads
Self-hosted databases create hidden costs for AI teams. Here's a real breakdown of what it actually costs to run your own PostgreSQL for vector search vs using a managed service.
7 min read