The founder

Gabriel Brien

AI Filmmaker

I direct AI films and ads, and I run the production myself, from the first idea to the final cut.

My short film Last One won the Grand Prize at AI Independent Montreal. This page is how I think about the craft, and why it shapes every project at Crimson Spark Agency.

Gabriel Brien, AI filmmaker and founder of Crimson Spark Agency
My vision

How I think about AI filmmaking.

Seven convictions that came out of three years of production experiments and thousands of hours spent testing what holds up on screen.

  1. 01

    Most weak AI videos don’t have a generation problem. They have a concept problem.

    A technically beautiful shot is still boring if nothing happens: no curiosity, no escalation, no surprise. I design the attention first and generate after. The question I ask on every project: what happens after the first impressive thing?

  2. 02

    Continuity is solved before the camera rolls.

    There is no set to go back to. Every character, wardrobe, prop and location is fixed on a reference before the first shot, so the same person walks from the first scene to the last, with the same jacket, the same scar, the same lollipop.

  3. 03

    I direct. I don’t just prompt.

    Camera position, geography, what the frame hides, when the reveal lands. I brief a shot the way I would brief a cinematographer: what happens first, what reacts, how it escalates, where it ends.

  4. 04

    Impossible, but believable.

    Realism has five layers: what you see, how things move and weigh, where a real camera could be, how people react, and how a real production would behave. Even when the event is impossible, it has to feel physically true.

  5. 05

    The workflow is the product, not the model.

    Models change every month. What lasts is the system around them: story, references, direction, edit, sound. Each step done with the right tool, and the final cut done by hand.

  6. 06

    AI didn’t make films easy. It moved the work.

    The hours moved from the set to the decisions: which take holds, which shot serves the story, what to cut. More control means more choices, and every one of them is a director’s choice.

  7. 07

    Finish the film.

    Production limits will force changes. I would rather simplify a scene than lose the film. The goal is the finished story, not loyalty to the first plan.

The path

From running operations to directing films.

Operations

Execution first

Years running operations, teams and launches. As COO at XPR Labs, I helped launch Taebo Reboot, a VR fitness game, and produced its official trailer.

Immersive tech

Spatial storytelling

Building VR and AR experiences taught me to think in space: where the viewer stands, what they see first, what the frame hides.

AI filmmaking

A new medium

Then generative AI became the medium. Three years of production experiments led to Crimson Spark Agency and to Last One.

Selected work

Stories I have directed.

Grand PrizeThe Independent Vision Award, AI Independent Montreal, for Last One
Official SelectionAi & Robots Film Festival 2026, Last One