Back to browse
GitHub Repository

Zero-config semantic search for any PostgreSQL database. pip install pgsemantic, three commands, done

48 starsPython

Postgres data cluster by meaning (semantic search and visualization)

by varmabudharaju·Mar 31, 2026·2 points·0 comments

AI Analysis

●●SolidBig BrainSolve My Problem

Database triggers keep embeddings in sync—no ETL pipelines or separate vector database needed.

Strengths
  • Trigger-based sync means embeddings update automatically as rows change
  • Local embedding model runs on your machine—no API keys or usage costs
  • Built-in visualization shows semantic clusters across your existing data
Weaknesses
  • Requires pgvector extension already installed on your PostgreSQL instance
  • Visualization is read-only—no way to interact with or export clusters
Category
Target Audience

Backend developers with existing PostgreSQL databases

Similar To

pgvector · LanceDB · Weaviate

Similar Projects

Developer Tools●●●Banger

PgCortex – AI enrichment per Postgres row, zero transaction blocking

Outbox + external worker architecture solves a real production pain: LLM calls in Postgres triggers without blocking transactions.

Big BrainSolve My Problem
supreeth_ravi
103mo ago
Developer Tools●●Solid

Diesel-guard adds custom checks via Rhai for Postgres migrations

It flags concrete dangerous patterns (example: ADD COLUMN with DEFAULT) and prints step-by-step safe alternatives instead of vague warnings. The embedded Rhai-based custom checks are a smart move — teams can encode org policies — but usefulness will depend on CI integration and the breadth/precision of its rule set.

Niche GemSolve My Problem
ayarotsky
103mo ago