LLMOps: Custom MCP Streamable HTTP connector for PostgreSQL database
Автор: The Machine Learning Engineer
Загружено: 2026-01-11
Просмотров: 56
In this video, I'll show you how to create a PostgreSQL MCP Server that uses the transport protocol (Streamable HTTP) and a Agentic Ai bot which use LangGraph, with a MCP client, and Streamlit as a front end.
Quick Architecture
**Database Layer**: `db.py` handles async connections with `asyncpg`, pooling, and metadata operations.
**MCP Server**: `server4.py` exposes tools over HTTP (including SSE transport and streamable HTTP) and a health check.
**Agents**: `agent_langchain.py` (direct) and `agent_mcp_client.py` (MCP client) package tools in LangGraph.
**UI**: `streamlit_chat.py` offers a modern chat interface with two agents, timers, export, and tool output handling.
Notebook and code are only available to paid subscribers. To access, contact the administrato, mlengineerchannel@gmail.com
https://github.com/olonok69/postgres_...
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке:
