Project / self-hosting

Self-Hosted Media Platform

A private media workflow that combines local storage, automated ingestion, and manual curation into one usable library.

Jun 24, 2026Status: archived

  • Linux
  • Docker
  • Jellyfin
  • Automation
  • Shell scripting

This project is a private media system, written up here as a public case study in workflow design.

It is less about a single app and more about a stack of decisions: where media comes from, how it is cleaned up, how it becomes visible to a library app, and how the parts stay separated so the public-facing experience remains stable.

What it is

A self-hosted media platform that combines three sources:

  • local media from a personal archive
  • automated media ingestion
  • manual link intake for curated library items

The public-facing goal is simple: keep browsing pleasant, even when the backend is doing messy work.

What it does

  • presents a single library view to the user
  • keeps raw backend storage out of the library surface
  • supports both automated and manual intake paths
  • filters noisy extras so the library stays readable
  • triggers refreshes so new items show up in the front end

How it works

  1. Local storage provides the personal archive when it is available.
  2. Automated ingestion handles the clean, routine media path.
  3. Manual intake handles approved links that need human curation.
  4. A cleanup step separates useful items from extras and other clutter.
  5. The library app reads only the curated paths.

That separation is the main design choice. It keeps the visible library simple while leaving the backend flexible.

Why it exists

The purpose is reliability and control.

Rather than exposing every source directly, the system narrows the user experience to the media that actually belongs in the library. That makes it easier to keep the collection usable, maintainable, and predictable over time.

What is intentionally omitted

  • public access to internal backend services
  • raw mount points or debug storage
  • credentials
  • secret files
  • private network topology
  • operator-only paths and ports

Technical themes

  • self-hosting
  • library curation
  • automation
  • cleanup and filtering
  • separation of concerns
  • practical media workflow design

Project note

The original snapshot included infrastructure details that are better kept private. This page preserves the concept, the workflow, and the purpose without leaking the operator layer behind it.

Further reading