We need the creative who is already pushing tools like Runway Gen-3, Kling, and Flux to their breaking points. The person who doesn't just type a prompt, but conducts an orchestra of parameters, seeds, camera movements, and compositing workflows to create something that feels human.

Who You Are (The DNA)

You are a filmmaker, but your camera is a neural network.

• You are a Tool-Hunter:

  You don't wait for tutorials. You are the first to test a new model on Hugging Face or replicate a workflow you saw on X (Twitter).

• You are Obsessed with Control:

  You hate "slot machine" AI generations. You use ControlNets, IP-Adapters, and precise motion brushes to force the AI to do exactly what you envisioned.

• You Have "The Eye":

  You understand cinematography—focal lengths, lighting ratios, composition, and cutting on the beat. You know why

  a shot feels emotional.

What You Will Actually Do

• World Building:

  Take a raw concept (a founder’s story, a spiritual epic, a futuristic ad) and define the visual physics, color palette, and atmosphere using Midjourney/Flux/VEO3

• Directing the AI:

  Generate high-fidelity video sequences. You will iterate 5 times on a single 3-second clip until the physics, lighting, and movement are indistinguishable from reality.

• Solving "The AI Problems":

  You will figure out how to keep a character's face consistent across 10 different shots. You will fix morphing hands. You will stabilize jittery backgrounds.

• Post-Production Alchemy:

  You won't just generate; you will composite. You’ll use After Effects/Premiere to stitch, retouch, and grade your AI footage into a cohesive narrative.

• R&D:

  You will spend 20% of your week just testing new tools (Veo, Sora, Hailuo, ComfyUI workflows) to ensure our studio remains state-of-the-art.

The Stack (If you know, you know)

We don’t expect you to know all of these, but we expect you to be dying to learn them:

• Video:

  VEO3, SORA2, Runway (Gen-3), Kling, Luma Dream Machine, Hailuo MiniMax, Pika.

• Image:

  Midjourney v6.1, Flux (Pro/Dev), Mystic, Nano Banana Pro etc.

• Advanced Control:

  ComfyUI (Nodes/Workflows), Automatic1111, LoRA training.

• Post:

  Topaz Video AI (Upscaling), DaVinci Resolve/Premiere, After Effects.

This Role is NOT for you if:

• You think AI video is just "typing a sentence and hitting enter."
• You get frustrated when the AI hallucinates; we need people who enjoy the puzzle of fixing it.
• You don't care about storytelling or pacing (pretty visuals without a story are useless to us).