# Valstorm Dedicated Video Worker (valstorm-video-worker)

> Dedicated asynchronous microservice for multi-preset FFmpeg video transcoding, Deepgram/Whisper speech-to-text captions, and AI chapter generation.

- **GitHub Repository:** https://github.com/Valstorm-Corp/valstorm-video-worker
- **License:** PolyForm Noncommercial License 1.0.0 (https://polyformproject.org/licenses/noncommercial/1.0.0)
- **Installation:** `git clone https://github.com/Valstorm-Corp/valstorm-video-worker.git && cd valstorm-video-worker && uv sync`
- **Language/Stack:** Python 3.12+ (FFmpeg, Deepgram/Whisper, Redis, valstorm-client)
- **Web Documentation:** https://www.valstorm.com/free-tools/valstorm-video-worker

## Key Capabilities
- **Resilient Multi-Preset FFmpeg Engine:** Supports stream-copy remuxing (fast_copy), balanced H.264/AAC encoding, and high-compression presets with automatic container fallback.
- **Automated Transcription & Captioning:** Extracts clean audio channels and generates word-level transcripts and VTT/SRT captions using Deepgram Nova-3 or local Whisper.
- **AI Chapter & Metadata Synthesis:** Turns raw video recordings and VSL audits into structured chapters, executive summaries, and searchable CRM records.
- **1:1 Valstorm Platform Synchronization:** Updates video status, transcript fields, and VFS file attachments automatically via the official valstorm-client Python SDK.

## Quickstart
```bash
# 1. Clone the repository and sync dependencies via uv
git clone https://github.com/Valstorm-Corp/valstorm-video-worker.git
cd valstorm-video-worker
uv sync

# 2. Start the Redis queue consumer and port 8080 /health server
uv run python -m main
```