← Back to Projects

AI-Assisted Production

Video Production Pipeline

An OBS-based capture rig with PTZ camera control and live Whisper captioning, paired with GPU-accelerated voice and music generation on RunPod so a single person can run a full production.

The capture layer runs on OBS Studio with PipeWire-based audio routing, PTZ camera control, and live Whisper-powered captions — no separate camera operator or captioner needed. On top of that, an AI generation layer offloads voiceover and music/SFX work to RunPod GPU workers, and an orchestration layer runs automated preflight checks before every session so audio/video drift gets caught before it ever hits a recording.

Pipeline layers

OBS Capture + PTZ + Captions

PipeWire-based audio capture, PTZ camera control, and live Whisper-generated captions during recording.

Live

AI Voice Cloning

GPU-accelerated text-to-speech and voice cloning (XTTS-v2) for narration and voiceover, run on RunPod.

In Development

Music & SFX Generation

GPU-accelerated background music and sound-effect generation, run on RunPod alongside the voice pipeline.

In Development

Studio Orchestration

Automated preflight checks — WiFi band, mic connection, audio filter chain — that catch AV drift before a session starts.

Live