setup.md

Markflare Setup

Requirements

• Clerk for authentication
• Supabase Project for database and vector search
• Cloudflare Account (Optional) for custom domain and AI Gateway(Monitoring)
• Vercel Account for deployment, CI/CD and serverless functions
• OpenAI API Key for AI API calls

Setup

Convert the .env.example file to .env.local and fill in the required values.

1. Create a Supabase Project

MarkAI uses Supabase for database and vector search. Create a Supabase project and copy the URL and API Key. Paste secrets in .env.local make sure they are correctly formatted. Copy SQL from tables.sql and run it in the Supabase SQL editor. This will create the required tables and functions.

2. Create a Clerk Application

MarkAI uses Clerk for authentication. Create a Clerk application and copy the API Key. Paste secrets in .env.local make sure they are correctly formatted. After creating the application, go to this docs page and follow the instructions to setup the application. Don't forget to activate organizatiions and teams in the Clerk dashboard.

3. OpenAI API Key

MarkAI uses OpenAI for AI API calls. Create an OpenAI account and copy the API Key. Paste secrets in .env.local make sure they are correctly formatted.

Midpoint Check

After last step MarkAI should be able to run locally. Run npm run dev and check if the application is running. If it is not, check if all the secrets are correctly formatted and saved in .env.local.

4. Deploy to Vercel

MarkAI uses Vercel for deployment, CI/CD and serverless functions. Create a Vercel account and connect it to your GitHub account. Create a new project and connect it to your GitHub repository. Make sure the Vercel project is connected to the correct repository and branch. Add the secrets from .env.local to the Vercel project. Make sure they are correctly formatted. Deploy the project to Vercel.