FixMyCoffee

FixMyCoffee

Timeline: 2025
Role: Solo Developer
Status: Offline
ReactTypeScriptSupabaseOpenAITailwind CSSViteFull-Stack

I'm a big coffee and espresso fan. Home espresso is a hobby with a real learning curve. Tiny parameter changes in dose, grind or brew time can completely change what's in the cup. FixMyCoffee started as a way to scratch that itch: an AI barista assistant that listens to your shot parameters, analyzes what went wrong and tells you exactly how to adjust. It was also my first attempt at building a complete full-stack web application from the ground up.

FixMyCoffee landing page

The FixMyCoffee landing page

The AI Barista Agent

The core of the app is a shot-logging flow where users enter their espresso parameters: dose, yield, brew time and a taste balance between sour and bitter. The agent then uses GPT-4o-mini to generate concise, specific feedback, paired with an automatic star rating calculated from target extraction ratios. The result is delivered with a typewriter animation to keep the experience a little playful.

Step 1: entering shot parameters

Step 1: log your shot

Screenshot of the AI shot analysis agent

Step 2: AI feedback

How it works

Log your shot parameters → select your flavor notes → the AI agent analyzes the extraction and gives you direct, actionable advice in under two sentences.

Blog & Content System

Alongside the AI tool, I built a full blog system backed by Supabase. Articles are written in Markdown and stored in the database, with an admin editing template for managing content. The blog renderer handles a table of contents, key takeaways, related posts and SEO meta tags, all processed through a Supabase Edge Function. The original goal was to drive organic traffic through espresso-related content, though visitor numbers turned out to be modest.

What I Learned

This was my first completed full-stack project and it covered a lot of ground in a short time. From setting up Supabase with row-level security and Edge Functions, to integrating the OpenAI API, deploying on Netlify with static site generation and wiring up product analytics via PostHog.

  • Full-stack development with React, TypeScript and Supabase
  • Building and deploying Deno-based serverless Edge Functions
  • Integrating OpenAI's API for real-time user-facing feedback
  • Static site generation with Vite React SSG for SEO
  • Building a content management system with a Markdown-based blog editor
  • Product analytics setup and conversion tracking

Technologies

Technologies

ReactTypeScriptSupabaseOpenAI APITailwind CSSViteDenoshadcn/uiNetlifyPostHog