Hello there 👋
I'm Mark Hou
Software engineer building thoughtful, well-designed systems.
About Me
I’m a software engineer with a B.Sc. in Computer Science at the University of British Columbia. I enjoy building tools that solve real problems, and I approach each project as an opportunity to improve how I write, structure, and reason about code.
As a new graduate, I’m continuing to explore different areas of software engineering — from application development to cloud infrastructure and distributed systems. I’m especially interested in how large-scale systems are designed and operated in production, and how design tradeoffs shape reliability and performance.
When I'm not working on a project, you'll find me out on a run, reading, or golfing badly.
Technologies I work with
TypeScript
React
Node.js
Python
Java
Postgres
aws
Git
Featured Project
ReviewFlow — Automated Code Review Routing & Analytics Platform
ReviewFlow is a full-stack SaaS application that streamlines the code review process for software development teams. As teams grow and pull requests accumulate, code reviews become a bottleneck, PRs sit unassigned, reviewers get overloaded, and development slows down. ReviewFlow solves this by automatically analyzing every pull request, scoring its complexity, and routing it to the most qualified available reviewer.
When a developer opens a pull request on GitHub, ReviewFlow instantly receives the event through a webhook, analyzes the changed files, calculates a complexity score based on lines changed, file types, and critical path detection, and assigns the best reviewer based on their file expertise, domain knowledge, and current workload. The entire pipeline, from PR opened to reviewer assigned, completes in under one second.
The platform features a real-time dashboard where team leads can monitor the review queue, track workload distribution across the team, and identify bottlenecks at a glance.
Key Features
- â–¸Automatic PR complexity scoring and priority classification (critical / high / medium / low)
- â–¸Intelligent reviewer assignment using a weighted SQL matching algorithm across file expertise, tag matching, and workload balancing
- â–¸Real-time dashboard with PR queue, complexity indicators, and team capacity overview
- â–¸GitHub OAuth login with team onboarding and multi-repository support
- â–¸Secure session-based authentication with team-scoped data access
Technical Highlights
- â–¸Serverless architecture on AWS (Lambda, API Gateway, S3, CloudFront) with Neon PostgreSQL
- â–¸GitHub webhook pipeline processing PR events end-to-end in under 1 second
- â–¸React + Tailwind CSS frontend deployed to a private S3 bucket behind CloudFront
- â–¸Privacy-preserving structured logging with automatic sensitive data redaction across all Lambda functions