JohnTech
EN 繁中

Selected work / Full-stack development

Ideas turned into
working products.

Interfaces people can use. Data models that support real workflows. Cloud infrastructure that brings it all together. Two projects I built across the stack.

Explore the projects
01

Product & interface

02

Backend & integrations

03

Cloud & delivery

01 / Property operations / Full-stack SaaS

StaySweep

A clearer handover
between every stay.

StaySweep public homepage interface
Public homepage · staysweep.net

The problem

Short-term rental turnovers involve properties, people, checklists, and tight schedules. Coordinating them across messages makes it hard to see what is assigned, what is done, and what still needs attention.

What I built

A connected workflow for hosts and cleaners: manage properties, create task templates, schedule cleaning batches, assign work, and review completion photos. Messaging and subscription billing support the wider product.

My contribution

Full-stack application development, database and authentication integration, APIs, media workflows, and AWS infrastructure as code.

01

Operations interface

Role-based host and cleaner areas, property management, reusable checklists, scheduling, and assignment status.

02

Connected backend

Supabase authentication and PostgreSQL data, signed S3 media workflows, SES invitations, Google Maps integration, and Lemon Squeezy billing webhooks.

03

Real-time & infrastructure

WebSocket messaging with API Gateway, Lambda, and DynamoDB, alongside Terraform-managed AWS services and CloudWatch monitoring.

Product workflow

  1. Property
  2. Schedule
  3. Assign
  4. Verify

Technology stack

  • Next.js
  • React
  • TypeScript
  • Tailwind CSS
  • Supabase / PostgreSQL
  • AWS S3 & SES
  • API Gateway & Lambda
  • DynamoDB
  • Terraform
  • Google Maps
  • Lemon Squeezy
Visit StaySweep

02 / AI application / Personal project

Harmony Healer

Thoughtful software
for better conversations.

Harmony Healer public homepage interface
Homepage preview · local project

The problem

A helpful conversation can be easy to forget. Reflection tools need a way to connect what someone shares with the small actions, habits, and notes they want to return to later.

What I built

An application that connects AI conversations with journals, mood and habit tracking, and follow-up plans. Shared and private context help organize the experience for individuals and couples.

My contribution

Full-stack product development, bilingual UI, authentication, relational data design, server-side AI integration, and context and memory workflows.

01

Bilingual product experience

English and Chinese interfaces for guided conversations, journals, habits, moods, and shared or private dashboard items.

02

Authentication & data

Supabase Auth with Google and Microsoft sign-in support, PostgreSQL migrations, and database access policies for user and couple data.

03

AI orchestration

Server-side OpenAI and Anthropic integrations, contextual prompts, memory summaries, suggested profile actions, and monthly message limits.

Product workflow

  1. Reflect
  2. Talk
  3. Plan
  4. Practice

Technology stack

  • Next.js
  • React
  • TypeScript
  • Tailwind CSS
  • Supabase Auth
  • PostgreSQL
  • OpenAI API
  • Anthropic SDK
  • SQL migrations

Your next project

Have a product in mind?

Let’s talk about the experience, the workflows, and what it will take to build it.

Discuss your project