Skip to content
Lu Phone Maw

Book Publishing Platform

A publishing management platform built around a real publisher's operation, bringing manuscript submission, editorial review, price negotiation, publication and reader discovery into one role-aware workflow.

Focus
Full-Stack, Client Project
Core stack
Next.js, TypeScript, Supabase/pgvector, Hugging Face, n8n, AWS (EC2)
Context
Built for Myat Pwint Book Publishing House
Links
Private client repository
Myat Pwint Books catalogue home page: a 'Describe what you want to read' search box with category, price and author filters above a grid of book covers.

How it works

Workflow diagram: authors submit a manuscript to the editor; the editor approves it to the publisher or rejects it for resubmission; the publisher negotiates price with the author and the book is published.

Manuscript workflow

Authors submit a manuscript to an editor, who approves it or sends it back for resubmission. The publisher then agrees the price with the author and publishes the book.

For readers

Myat Pwint Books catalogue with the search 'books about starting a business' and AI search active, showing business-related books such as The Entrepreneur's Mindset and Financial Freedom Blueprint.

AI search

Readers describe what they want — here, “books about starting a business” — and get books matched by meaning, not just by title.

Book page for 'Whispers of the Irrawaddy' by Lu Maw, showing the cover, genre tags, an About This Book summary, and published date, edition and language.

Book page

Each book has its own page with the cover, author, genres and summary, plus the publication date, edition and language.

Checkout page with one digital book in the order, an order summary of 25,000 MMK, and payment options for Stripe and a demo payment for testing.

Checkout

Readers buy straight from a book’s page. This build offers a Stripe payment option and a demo payment mode used for testing.

For editors and publishers

Editor dashboard with counts for all, pending, under review, approved, rejected and published books, and a manuscripts queue with cards showing a resubmission badge, View DOCX and Feedback buttons.

Editor dashboard

Manuscripts are tracked by status — pending, under review, approved, rejected, published. Each one has its DOCX, a feedback thread and a resubmission count.

Publisher dashboard with a sales and revenue overview (revenue, sales, books and customers), book counts by status, and approved manuscripts below.

Publisher dashboard

Sales and revenue at the top, then every approved manuscript waiting to be published.

Two manuscript cards in the publisher view: one approved and ready to publish with a Publish button, and one published with price, physical inventory of 500 printed, and sales performance.

Publishing a book

An approved manuscript gets a Publish button. Once it’s published, the card tracks the price, the physical print run and stock, and sales.

Marketing automation with n8n

'Publish once. AI markets everywhere using n8n': publishing a book triggers a webhook, a content-generation AI agent drafts content, a human reviews it with a feedback loop back to the agent, and approved content goes to Telegram, email, Instagram and Facebook. An n8n workflow canvas is shown alongside.

Publish once, market everywhere

Publishing a book triggers an n8n workflow through a webhook, and an AI agent drafts the marketing content. A person reviews every draft — sending feedback back to the agent, or approving it — and approved posts go out to Telegram, email, Instagram and Facebook.

System architecture

Overview system architecture: AWS Application Load Balancer, then PM2 process manager, then a Next.js app containing the React/TypeScript frontend and API-route backend, then Supabase Cloud, then external integrations.

Overview

Traffic comes in through an AWS Application Load Balancer to one Next.js app, kept running by PM2, that holds both the React frontend and the API routes. The app talks to Supabase Cloud, which sits in front of the external integrations.

Three-tier system architecture: deployment (Cloudflare, AWS load balancer, EC2 running Next.js), application layer (React 19 frontend and API routes), data layer (Supabase Postgres with row-level security, Auth, Storage and Realtime), and external services (OnlyOffice, KBZPay, Hugging Face embeddings, n8n, Resend).

Three-tier architecture

Cloudflare and an AWS load balancer sit in front of the Next.js app on EC2. Its API routes handle manuscript uploads, approvals, OnlyOffice editing, AI embeddings and checkout. Supabase provides Postgres with row-level security, auth, file storage and realtime chat; external services cover document editing, payments, embeddings, marketing and email.

All projects