Investigations

VideoVox

Go June 2025

A video processing suite with subtitle generation, AI voice-over, and silence removal — plus a SaaS-ready Next.js frontend.

VideoVox combines three Go CLI tools I kept needing into one unified system. freeze-cut removes silence and freeze frames from recordings. voice-over generates AI-enhanced text-to-speech with multiple providers (OpenAI, Google, ElevenLabs, Gemini 2.5 Pro). caption-gen handles subtitle generation with automatic translation.

An orchestrator ties them together so you can pipe a raw recording through the full pipeline. The system supports multi-language translation (Portuguese, Spanish, French to English) and time-range processing for targeting specific segments.

The SaaS architecture adds a Next.js 15 frontend, hexagonal backend with event sourcing, and a 12-container Docker setup with monitoring. CI/CD runs through GitHub Actions with 7 quality jobs and Codecov integration maintaining 80%+ backend test coverage.

There's even a mobile app in the works. This project evolved from earlier experiments (subtitler, freeze-cut) into something genuinely production-ready.

What this project demonstrates

Three integrated CLI tools: freeze-cut, voice-over, caption-gen
Multi-provider TTS: OpenAI, Google, ElevenLabs, Gemini 2.5 Pro
Automatic multi-language translation
SaaS architecture with hexagonal design and event sourcing
12-container Docker setup with monitoring
CI/CD with 7 quality jobs and 80%+ test coverage

Technologies

Video ProcessingAITTSDockerNext.jsEvent Sourcing

© 2026 Jonathan Leahy · v0.8.3