CreatorOS — Gameplay Video Pipeline
Internal video automation tool actively used for a Bavolar Studios YouTube channel, built around a configurable Python and FFmpeg pipeline.

Review library in the studio's prototype interface: video packages waiting for my review, on my own machine with my channel's recordings.
Project overview
CreatorOS is an internal Bavolar Studios tool that I actively use for my YouTube channel's gameplay-video workflow. Its current implementation is a configurable video-processing pipeline, with ongoing development guided by that use.
Bavolar Studios brings together my YouTube channels, freelance development, and work on games, apps, and services.
Engineering approach
- Python CLI and YAML profiles configure processing stages for different editing workflows.
- Staged processing separates ingest, audio cleanup, transcription, clip selection, timeline assembly, and rendering.
- FFmpeg handles media processing; optional Whisper integrations support transcription.
- Per-job workspaces organize intermediate artifacts and logs by stage, making individual runs easier to inspect.
- Optional narrative assistance and review workflows keep creator decisions part of the editing process.
Studio interface
A local web interface, still a prototype, sits on top of the pipeline. It has a review library for finished video packages, a map of the pipeline's steps showing which are automatic and which wait for me, and an upload calendar. It runs on my own machine for my channel.
Current use
The tool is in active internal use for my channel while development continues. That workflow helps guide improvements to the editing process.
Future direction
I want to develop CreatorOS into a monetized product. The longer-term goal is to help creators turn gameplay or livestream recordings into videos with the quality and publishing frequency associated with a production team. That is a product goal, not a demonstrated outcome of the current pipeline.
Project Details
- Role
- Developer / Creator
- Timeline
- Ongoing
- Status
- In Use / Active Development
- Technologies
- PythonFFmpegYAMLPydanticWhisper (optional)