This podcast episode delves into the intricate process of building an AI fitness coach application using NextJS, OpenAI, and various complementary technologies, providing a comprehensive guide to setting up features such as user authentication, backend integration, and real-time communication, while also showcasing innovative tools for personalized fitness guidance.
Building an AI Fitness Coach with NextJS and OpenAI
Project Overview and Setup
Authentication with Clerk
Structuring the Application with Route Groups
Creating the App Layout with a Navbar
Building Goggins AI with OpenAI Assistants
Understanding Assistant Workflows
Implementing Backend Routes for Thread, Message, and Run
Using Postman to Test Backend Routes
Setting Up Planet Scale and Prisma for Database Integration
Building the Chat UI with State Management
Addressing Roadblocks and Introducing Global State with JOTA
Fetching Messages with Global State and Polling
Fetching and Storing Thread ID with User Authentication
Implementing a Backend Endpoint for User Threads
Adding a Chat Input and Send Functionality
Sending Messages and Triggering OpenAI Assistant Runs
Implementing Run Management and Polling
Styling Chat Messages Based on Sender
Deploying the Application to Vercel
Handling User Profiles and Preferences
Implementing Backend Routes for Challenge Preferences
Sending Workout Challenges to Users
Testing Workout Challenge Delivery with Postman
Setting Up Automatic Challenge Delivery with Upstash
Enabling Push Notifications and Client-Side Registration
Creating a Backend Endpoint for Saving Push Subscriptions
Sending Notifications with Web Push
Testing Push Notifications on a Phone
Sign in to continue reading, translating and more.
Open full episode in Podwise 
