💡 This document is a living record of how FishStudio was designed, what changed during implementation, and why those changes were made. Written for anyone who wants to understand the real engineering thought process — not the polished final result, but the messy, iterative truth.

📌 Table of Contents

1. The Spark — Why This Problem
2. V0 Blueprint — What I Originally Planned
3. Architecture: The First Sketch
4. Module-by-Module Design Thoughts
5. The Pivot Points — What Changed & Why
6. Security: How My Thinking Evolved
7. Data Layer Decisions — Why Polyglot
8. Async & Real-Time: The Hard Part
9. Deployment Evolution
10. Honest Retrospective

🌊 1. The Spark — Why This Problem

The Real-World Problem I Was Solving

Fresh fish in India is sold at wet markets. You go in person, haggle, carry it home. The entire supply chain is offline. There's no equivalent of BigBasket or Blinkit that handles fish well because:

• Fish is weight-based, not unit-based — a customer says "give me 500g of rohu," not "give me 1 rohu"
• Fish is ultra-perishable — it needs to move within hours, not days