Hyper-Realistic AI Video Series

I’m kicking off a character-driven, short-form video series and need an AI video specialist who can turn a single custom avatar into a lifelike recurring “host.” The look I’m after is the seamless, cinematic realism you see in today’s viral shorts—consistent facial detail, natural motion, dynamic environments, and smooth camera moves.

Tooling
I plan to work on a third-party generative video platform, but I haven’t locked one in. If you have proven workflows in DeepBrain, Synthesia, Runway ML, or another comparable solution, I’m ready to follow your lead on whatever gets us the best photorealistic results with the least friction.

What you’ll tackle
• Recreate my supplied character concept so the avatar can be reused across episodes without drift in appearance or quality.
• Build a repeatable pipeline for 30–60 second vertical videos covering rotating professional and creative topics.
• Deliver the first batch of clips (with audio, subtle motion graphics, and basic colour work) and document the process so we can scale future episodes quickly.

Acceptance markers
• Avatar match: near-identical to the reference in every frame.
• Output: 1080 × 1920, ready for TikTok/Reels, <60 s each, mastered with clean dialogue and light background music.
• Source project files and any prompt/parameter presets handed off for future use.

If you thrive on pushing hyper-realistic AI video, can advise on the best platform, and enjoy building repeatable systems, let’s get started.