Reddit Cluster Map is a full-stack data visualization project that crawls Reddit, models relationships between communities and users, and renders those connections as an interactive graph. As a portfolio piece, it highlights the kind of work I do professionally: backend systems, data pipelines, APIs, performance-minded frontend development, and production-ready tooling.
- End-to-end ownership — designed and built across backend, database, graph processing, frontend, and developer tooling
- Strong backend fundamentals — Go services for crawling, API delivery, graph precalculation, rate limiting, retries, and scheduled jobs
- Data and graph engineering — PostgreSQL-backed normalized storage with precomputed graph data and community detection
- Interactive frontend — React + TypeScript UI with 2D/3D network visualization, dashboards, and inspection workflows
- Production mindset — Dockerized local stack, CI workflows, Prometheus/Grafana monitoring, performance testing, and security-focused docs
- Backend: Go, PostgreSQL, sqlc, REST APIs
- Frontend: React, TypeScript, Vite, D3, Three.js
- Infrastructure & quality: Docker, GitHub Actions, Prometheus, Grafana, Vitest, Playwright
This project highlights my ability to:
- build and structure multi-service applications
- work comfortably across APIs, databases, and frontend visualization
- design for reliability with retries, rate limiting, monitoring, and testing
- turn complex data into an experience that is useful, performant, and easy to explore
- Architecture and system overview:
docs/overview.md - API documentation:
docs/api.md - Setup and local development:
docs/setup.md - Developer workflows:
Makefile
If you're reviewing this as part of my portfolio, the main takeaway is simple: Reddit Cluster Map reflects full-stack product engineering with an emphasis on data-heavy systems, clear architecture, and polished execution.