Ewen Chou ewenchou

LLM Wiki

A pattern for building personal knowledge bases using LLMs.

This is an idea file, it is designed to be copy pasted to your own LLM Agent (e.g. OpenAI Codex, Claude Code, OpenCode / Pi, or etc.). Its goal is to communicate the high level idea, but your agent will build out the specifics in collaboration with you.

The core idea

Most people's experience with LLMs and documents looks like RAG: you upload a collection of files, the LLM retrieves relevant chunks at query time, and generates an answer. This works, but the LLM is rediscovering knowledge from scratch on every question. There's no accumulation. Ask a subtle question that requires synthesizing five documents, and the LLM has to find and piece together the relevant fragments every time. Nothing is built up. NotebookLM, ChatGPT file uploads, and most RAG systems work this way.

Install dlib and face_recognition on a Raspberry Pi

Instructions tested with a Raspberry Pi 2 with an 8GB memory card. Probably also works fine on a Raspberry Pi 3.

Steps

Download the latest Raspbian Jessie Light image. Earlier versions of Raspbian won't work.

Write it to a memory card using Etcher, put the memory card in the RPi and boot it up.

	<!DOCTYPE html>
	<html>
	<head>
	<script src="http://code.jquery.com/jquery.min.js"></script>
	<link href="http://twitter.github.com/bootstrap/assets/css/bootstrap.css" rel="stylesheet" type="text/css" />
	<script src="http://twitter.github.com/bootstrap/assets/js/bootstrap.js"></script>
	<meta charset=utf-8 />
	</head>
	<body>

	var demoApp = angular.module('demoApp', ['ngResource'], function($locationProvider) {
	$locationProvider.hashPrefix('');
	});

	function MainCtrl($scope, Serv) {
	$scope.selectedItem = {
	value: 0,
	label: ''
	};
	$scope.Wrapper = Serv;

	//
	// This server will start a bash shell and expose it
	// over socket.io to a browser. See ./term.html for the
	// client side.
	//
	// You should probably:
	//
	// npm install socket.io
	// curl -O https://github.com/LearnBoost/Socket.IO/raw/master/socket.io.min.js
	//