Human Synergy + Custom AI: Delivering Under Pressure
- Jun 2
- 2 min read
We recently faced a high-stakes delivery challenge: leading the production of 35 critical message videos for a major global event on an incredibly tight turnaround. The assets were arriving from all over the world, featuring a diverse mix of high-profile contributors, from CEOs and sports stars to celebrities. With global stakeholders and sensitive content, I had to ensure two things: absolute data security and pristine dual-language subtitling adapted to a specific, event-themed format. Standard out-of-the-box software like Premiere Pro, of course, has built-in transcription and translation, but I’ve never found it precise enough in terms of nuance, formatting, and accurate Japanese translation; therefore, a more refined approach was needed. The solution was to upgrade my custom-built Subtitle Editor tool with targeted AI features. This included transcriptions via OpenAI’s Whisper model and multiple language translations by GPT-5, along with the event-specific subtitle format built into the AI instructions with a one-click auto-format button. Since these were high-profile message videos, the tool was routed through a closed-loop API, providing the speed and security needed for the job. By designing and introducing this innovative workflow, we moved the subtitling to the beginning of the process, allowing us to hand off the edits to global stakeholders for rapid sign-off as quickly as possible and keeping us on a tight schedule. The tool has helped us cut our subtitling workflow by 50% compared to our previous process. But as much as I love building these custom workflow solutions, tech is only half the equation. What really made this project a success was also the human element. The team remained incredibly organized, strictly focused on their objectives, and executed flawless, rapid hand-offs to keep the momentum going. Pairing a highly coordinated human team with a tailored, secure AI tool is where the real magic is beginning to happen.
Below is a gallery of the Subtle Editor App used for this project.

