Creative X had an adoption problem with three parts. Everyone needed to get up to speed on the creative tools that would change how we do our jobs. We needed to solve high-friction pain points. And where no existing tool could solve them, we needed to build our own.
What I built
As the creative AI lead across Creative X, I put the infrastructure in place to solve all three.
- Third-party tool evaluation, testing, and onboarding. One process covering contracting, team-wide tool spend, which pain points each tool solves, and the onboarding and training that make people actually adopt it. Includes vendor pilots like LTX Studio.
- Producer AI Playbooks. Guides that show producers across the department how to use AI, from agents to tools like NotebookLM, to make their own jobs easier.
- First-party tools. Where third-party tools fell short, I made the case for building our own and got engineering support:
- Automated asset resizing for layered files, built into the localization pipeline.
- AI-assisted localization pipeline. Transcreation with AI, including ElevenLabs dubbing, which cuts agency spend and moves faster.
- Product accuracy pipelines that keep Reality Labs' physical products accurate in AI-assisted creative.
[Diagram: evaluate, pilot, approve, train, scale]
Scale & Impact
- Department-wide across Creative X: 400 people in 15 markets
- 80% of the team trained; 15 tools evaluated, 8 vetted, 3 onboarded
- An estimated multimillion-dollar savings opportunity year over year from resizing, versioning, and localization
- First-party tools that are Creative X's own IP, strengthening its agency offering within Meta and beyond