May 28, 2026
Building a personal information system
The first version favors static publishing, a small API surface, and Cloudflare free-tier primitives.
- cloudflare
- architecture
- personal-systems
The system starts with a boring rule: public content belongs in Git, while private or dynamic data belongs behind an API.
That split keeps the website fast and resilient. The resume and blog can be served as static files, so they do not depend on a database or a warm backend. The API can stay small: health checks, metadata, and document CRUD over KV.
Why KV first
Workers KV is a pragmatic document store for personal data that is read often and
written occasionally. It is not a relational database, and it should not pretend
to be one. The v1 API treats KV values as JSON documents and keeps key naming
explicit: docs/{collection}/{id}.
What waits
File uploads, R2-backed media, search, and a private admin UI can arrive later. The first version is intentionally plain enough to deploy, inspect, and replace.