How I think

Four principles I keep coming back to. They show up in every project I take on.

AI amplifies, it doesn’t replace

I’m not building autonomous systems that try to make creative decisions. I’m building tools that handle the tedious parts — transcription, organization, catalogs, asset routing — so the human stays in charge of taste and judgment. The best AI in production is the kind you forget is there.

Stories transport, they don’t lecture

The most common mistake in nonprofit and ministry video is telling the audience what to feel. The audience doesn’t want to be lectured. They want to be a witness. I aim for stories that let people encounter something true, then let them draw their own conclusions.

Domain expertise + AI beats either alone

The best AI tools I’ve built came from knowing exactly what was broken about the existing workflow — not from being a great engineer. A coder without the domain ends up building the wrong thing well. A director without the code can’t build at all. The combination is what makes it useful.

Ship it

I’d rather have a rough system running in production for six months than a polished prototype that never deploys. Real systems get used; they get feedback; they evolve. Prototypes don’t. Every agent I’ve built has been deployed and is running right now.

On faith and work

I’m a Christian. The reason I lean toward ministries and small nonprofits is that I think professional storytelling should be accessible to organizations that can’t normally afford it. The work and the faith are the same thing for me — I don’t separate them. But I don’t preach through the work, either. I tell stories.

On where I come from

I grew up in Bogotá. I moved to the United States in 2007. I think in two languages, which is most of the time an advantage, occasionally a tax. My instinct as a storyteller comes from there — Colombia teaches you that ordinary lives, told well, are the most worth telling.