Preview
Open Original
Last week, I was a LAMP stack developer. This week, I’m shipping AI apps on Vercel. I built GitViral—a tool that converts technical READMEs into marketing content. Here is the technical breakdown of how I did it with a $0 budget:
- The AI Engine (Groq + Llama 3.3) I chose Groq over OpenAI because of the latency. When you’re building a "content generator," users want instant results. As you can see in my API route, the implementation is a simple POST request using the Groq SDK.
- The Database (Supabase) Coming from MySQL, Supabase felt like a superpower. I used their "Magic Link" auth and a simple profiles table to manage user credits. I even set up a PostgreSQL trigger to give new users 3 credits automatically upon sign-up.
- The Frontend (Next.js + Tailwind) Swi…
Last week, I was a LAMP stack developer. This week, I’m shipping AI apps on Vercel. I built GitViral—a tool that converts technical READMEs into marketing content. Here is the technical breakdown of how I did it with a $0 budget:
- The AI Engine (Groq + Llama 3.3) I chose Groq over OpenAI because of the latency. When you’re building a "content generator," users want instant results. As you can see in my API route, the implementation is a simple POST request using the Groq SDK.
- The Database (Supabase) Coming from MySQL, Supabase felt like a superpower. I used their "Magic Link" auth and a simple profiles table to manage user credits. I even set up a PostgreSQL trigger to give new users 3 credits automatically upon sign-up.
- The Frontend (Next.js + Tailwind) Switching from Bootstrap to Tailwind was the best decision for UI speed. The Result: A fully functional, authenticated SaaS in 2 days. You can see it live here: [https://git-viral.vercel.app/] Challenge: I’m currently looking for my first 10 "Prosumer" users. If you have a GitHub project that needs more eyes on it, try running your README through it and let me know if the "Technical Thread" it generates is actually accurate.