Uploaded April 2025 | Updated September 2026, 2 weeks ago

---

In this project, we will build a full stack app that works similar to ChatGPT and answers prompts and keeps the context throughout the conversation. We will use Vue.js with Pinia for state management, Node and Express on the backend with TypeScript. Stream for the chat featrues and a Neon PostgreSQL database to store users and chat logs.

Stream Chat:
getstream.io/chat/sdk/react/?utm_source=youtube&utm_medium=referral&utm_content=&utm_campaign=traversymedia

Neon Database:
https://fyi.neon.tech/1traversy

Code (Node/Express Backend):
github.com/bradtraversy/chat-ai-api

Code (Vue.js Frontend):
github.com/bradtraversy/chat-ai-ui

Timestamps:
0:00 - Intro
2:43 - Documentation URLs
4:13 - Backend Init & Install Dependencies
7:56 - Backend TypeScript Config
10:03 - Express Server Setup
13:35 - /register-user Endpoint Setup
17:37 - Initialize Stream With API Keys
21:17 - Register User With Stream Chat
27:41 - /chat Endpoint & Open AI Init
42:49 - Neon PostgreSQL Database Setup
46:00 - Database Config & Neon Adapter
48:46 - Drizzle Schema
54:26 - Drizzle Config
57:31 - Database Migration
59:11 - Save Users In Neon PostgreSQL
1:04:58 - Save Chats In Neon PostgreSQL
1:09:38 - Get User Chat History
1:15:22 - Frontend Init & Install Dependencies
1:19:21 - TailwindCSS Setup & Clean Up
1:22:02 - Router & Page Setup
1:27:54 - Home Form Display
1:31:52 - Reactive Variables & Input Bind
1:37:24 - Pinia Initialization
1:39:19 - Pinia User Store
1:42:57 - Hook Up Home Form
1:50:02 - Header & Leave Chat/Logout
1:57:20 - Pinia Chat Store
2:01:45 - Load Chat History Function/Request
2:08:42 - Load Chat Into UI
2:17:50 - Send Message Function/Request
2:22:21 - Send Message Form
2:29:32 - Get Context Of Chat
2:37:37 - Format Output Text
2:40:52 - Backend API Deploy To Render
2:43:55 - Frontend Deploy To Vercel