PostgreSQL vs Supabase

Vector Database vs Knowledge Graph for LLMs

A practical guide comparing vector databases and knowledge graphs as storage backends for LLM applications. The article explains that vector databases (e.g., Qdrant, pgvector, Pinecone) store embeddings and excel at semantic similarity queries, while knowledge graphs (e.g., Neo4j, RDF stores) model entities and relationships and support exact, multi-hop relational queries. The author argues most real-world LLM systems need a hybrid: vectors for retrieval and graphs for precise reasoning and auditability (a pattern exemplified by Microsoft’s GraphRAG). Tool-level trade-offs, operational costs, and a four-question decision rule are provided to help teams choose the right architecture for their queries.

• Vector databases store high-dimensional embeddings and are optimized for semantic similarity (examples: Qdrant, pgvector, Pinecone, Milvus, Weaviate, Chroma).
• Knowledge graphs store entities and relationships as first-class citizens and support exact traversals and multi-hop queries using languages like Cypher or SPARQL (examples: Neo4j, RDF stores).
• The author has deployed both approaches in production: vector solutions (Qdrant and pgvector) and Neo4j for graph workloads.

pg_statviz 1.2 released with PostgreSQL 19 support and new features

Just in time for the PostgreSQL 19 betas, I'm excited to announce release 1.2 of pg_statviz, the minimalist extension and …

Ship a RAG Chatbot with Claude, pgvector, FastAPI

This technical how-to shows how to build a retrieval-augmented generation (RAG) chatbot in a weekend using three components: PostgreSQL with the pgvector extension as the vector store, FastAPI as a thin web layer, and Anthropic's Claude for text generation. Claude does not provide an embeddings endpoint, so an external embedding provider is required (examples used are Voyage AI, OpenAI, or local sentence-transformers). The guide explains practical schema choices (embedding dimension must match pgvector column), ingestion and chunking, HNSW indexing for fast nearest-neighbor search, and the single-SQL retrieval pattern (ORDER BY embedding <=> query LIMIT k). It also emphasizes grounding via a strict system prompt and lists production hardening tasks (connection pooling, evaluation, better chunking, streaming, and auth) that should follow the initial weekend prototype.

• Postgres with the pgvector extension can serve as a vector store; create a vector column sized to the embedding model and use an HNSW index for fast nearest-neighbor search.
• Anthropic's Claude (generation) does not provide an embeddings API; the author uses an external embedding provider (Voyage AI's voyage-3, OpenAI, or local models).
• Retrieval is implemented with a single SQL query using cosine distance: ORDER BY embedding <=> query LIMIT k to get top-k chunks.

Supabase is now available in Gemini Enterprise

Connect Supabase to Gemini Enterprise and query your Supabase projects with natural language, right inside the platform.

Build SaaS Apps with Claude Code, Clerk, and Supabase

This technical walkthrough guides product managers on how to build, secure, and monetize a full-stack B2B2C SaaS application using AI agents without writing code. Using a case study called AskOne—a live Q&A alternative to Slido—the tutorial details a universal, multi-step workflow. The process utilizes Anthropic's Claude Code to generate design layouts, establish design system tokens, and implement functional React components. The system integrates Clerk for user authentication and recurring subscription billing configuration, alongside Supabase as the backend database connected securely via Claude's Model Context Protocol (MCP) server framework.

• The author developed Grok Build for VS Code, achieving over 105,000 installations and 37,000 monthly active users (MAU).
• AI agents autonomously wrote 5,094 automated end-to-end integration and unit tests for the software in three months.
• The workflow utilizes a recommended B2B stack consisting of Next.js, Clerk, and Supabase.

Safely Drop All PostgreSQL Tables (2026)

A technical how-to describing safe methods to drop all tables in a PostgreSQL database. The article shows a one-command reset (DROP SCHEMA public CASCADE; CREATE SCHEMA public; GRANT ...), explains PostgreSQL 15 default changes that revoke CREATE from PUBLIC and set public's owner to pg_database_owner (which can cause permission errors), and details Supabase-specific risks (Dropping public can remove Supabase-managed schemas and extensions). It gives variants to preserve extensions (reinstall extensions or drop tables only via a DO-block loop), recommends wrapping resets in transactions for local dev, and provides a production safety checklist: take backups, confirm connection, block connections, and restore from pg_dump/pg_restore rather than dropping schema in prod.

• DROP SCHEMA public CASCADE removes the public schema and all dependent objects (tables, views, sequences, functions, triggers).
• PostgreSQL 15 defaults: PUBLIC no longer has CREATE by default, and the owner of public is pg_database_owner, which can cause 'permission denied for schema public' after recreating the schema.
• On Supabase, DROP SCHEMA public CASCADE can cascade into Supabase-managed schemas (auth, storage, realtime, graphql) and break the project; use 'supabase db reset' for dev and restore from backups for production.