Juan Ocampo

Juan
Ocampo

AI video director and photographer

Watch the work Instagram LinkedIn

About

For four years I worked at Dream But Don't Sleep, a creative studio in Jersey City, where I shot, edited, brainstormed and helped bring to life projects for Red Bull, Timberland, Coinbase, Adidas, Fortnite and many more. It showed me all the gears turning behind a finished piece that most people never see.

My background is in photography and videography, and what I bring to the table most is being a visionary, a connector and a creator.

Now I make AI films end to end for brands big and small, and help clients build out their media and social content with the tools that best suit their needs. I also build websites with AI, run OASYS Studios, host my podcast, Tribe, wrote a book called 22, and work in English and Spanish.

I'm always learning: I founded an AI brain trust where we trade ideas and tests on the newest tools, and I've read 100+ books across business, philosophy and AI. I believe AI isn't just the future, it's the present, and every business should be adapting to it now.

Tools I use

Video
Higgsfield, Kling, Google Flow, Veo, Runway, Sora, Seedance, Hailuo, Wan, Magnific, Weavy, HeyGen
Image
Midjourney, Nano Banana, Flux, Krea, Magnific, Adobe Firefly, ChatGPT
Sound
ElevenLabs, Suno, Artlist, Adobe Firefly
Writing
Claude, ChatGPT, Gemini
Edit
Premiere Pro, Lightroom

AI films

Brand pieces marked “Spec” are self-initiated, made to show what I would pitch to that brand.

Brand work

Some of the work I was part of while working with Dream But Don't Sleep.

Camera work

Before AI, I learned on a real camera. Events, behind-the-scenes coverage, real estate, documentary and portraits, all shot and edited by me. It's why my AI work is framed and lit like it was filmed.

Photography

Websites

When a project calls for it, I also design and build websites with AI.

Writing

Cover of 22: Thoughts by Juan Esteban Ocampo

22: Thoughts

My book of poetry and personal reflection on love, loss, self-discovery and purpose. 302 pages, published 2025.

Get it on Amazon

From idea to final cut

Every AI film above went through the same five stages. I can take a project through all of them myself, or own any single stage inside a larger team.

  1. Script and boards

    Find the one idea the film has to land, write it in beats with Claude, ChatGPT and Gemini, and board every shot before generating anything, using Adobe Firefly and Magnific for quick frames.

  2. Look development

    Lock characters, wardrobe, product and lighting as stills in Midjourney, Nano Banana, Flux and more, so the same character, the same product, or both show up the same in every shot.

  3. Motion

    Animate the approved frames in Kling, Veo, Seedance and more, directing camera moves shot by shot rather than hoping for a lucky take.

  4. Voice and sound

    Voiceover, music and effects with ElevenLabs, Suno, Artlist and Adobe Firefly, timed to the cut instead of pasted on at the end.

  5. Edit and finish

    Cut, grade and deliver in Premiere Pro, in whatever format the placement calls for.