What's in the pgEdge AI Toolkit?
Everything you need to build enterprise-grade AI-powered applications on PostgreSQL.
pgEdge Postgres MCP Server
Connect AI assistants directly to PostgreSQL
Built-in PostgreSQL knowledge base
Natural language SQL. Default read-only. Optional write mode
Works with any PostgreSQL 14+ (RDS, Aurora, etc.)
OpenAI, Voyage AI, or Ollama embeddings
pgEdge RAG Server
Build AI apps with semantic search and retrieval
Hybrid search — vector + BM25 + Reciprocal Rank Fusion
Streaming responses via Server-Sent Events
Multi-turn conversations with source attribution
LLM providers: OpenAI, Anthropic, Ollama
pgEdge Docloader
Load docs into PostgreSQL for vectorization
HTML, Markdown, reStructuredText, SGML
Clone and load directly from Git repos
Upsert mode for automated sync pipelines
Ready for automatic embedding
AI Extensions
PostgreSQL extensions for AI workloads
pgVector — HNSW & IVFFlat similarity search
pgedge_vectorizer — Automatic embedding generation
VectorChord BM25 — Native keyword ranking
pg_tokenizer — Text tokenization
pgEdge Anonymizer
Mask PII in database copies
105 built-in patterns — names, emails, SSNs, and more
JSON/JSONB column support
Preserves referential integrity and formatting
19 countries with locale-specific formats
See the MCP and RAG Server in Action
A complete, PostgreSQL-native stack for building production AI applications.
MCP Server — AI Agents Meet PostgreSQL
See how AI agents query, analyze, and optimize a PostgreSQL database through the MCP Server.
RAG Pipeline — Build AI Apps on Your Data
Ask Ellie is a live app built with the pgEdge AI Toolkit. No external vector databases required.
Try the pgEdge MCP Server yourself
Pick the path that works for you — from zero-install browser walkthroughs to running on your own machine.
Run it in Claude Code or Desktop
Connect to your own database, or try it with a sample Northwind DB.
Open Terminal (Mac: search Spotlight for "Terminal")
Paste this command and press Enter:
curl -fsSL https://raw.githubusercontent.com/pgEdge/pgedge-postgres-mcp/main/examples/claude-try-it/install.sh | bashFollow the prompts (~2 minutes)
Claude Code: start a new conversation. Claude Desktop: quit (Cmd+Q) and reopen.
Ask: "What tables are in my database?"
Docker Compose Demo
Launches a web-based chat agent — no AI coding tool needed. Just Docker.
/bin/sh -c "$(curl -fsSL https://downloads.pgedge.com/quickstart/mcp/pgedge-ait-demo.sh)"
Codespaces Walkthrough
Full VS Code environment with interactive run buttons — nothing to install.