About The Position

Join a leading AI lab’s Generative AI team to help develop next-generation video generation models focused on realistic human performances, cinematic storytelling, and accurate scene recreation. The role is intended for experienced film, television, or animation professionals with strong backgrounds in directing, performance, staging, and visual storytelling. You will train and evaluate advanced AI video models by assessing how well they capture and reproduce acting realism, character blocking, screen direction, eyelines, emotional expression, scene continuity, and cinematic composition. Your creative judgment will help establish evaluation benchmarks, identify model limitations, and improve the production of believable, emotionally compelling, professionally directed scenes. This is a full-time engagement of approximately 40 hours per week, with consistent weekday availability, as part of a leading AI research lab’s extended creative workforce. Compensation is $70–$110 per hour.

Requirements

  • At least 7 years of professional experience in film, television, theater, animation, or a related production environment.
  • Professional credits on at least 2 films or equivalent high-quality professional productions.
  • A strong professional background in at least one relevant role: Film Director, Television Director, First Assistant Director (1st AD), Animation Director, Character Animation Supervisor, Script Supervisor, Performance Capture Director, or Acting Teacher/Acting Coach with professional film or theater experience.
  • Deep understanding of filmmaking principles, cinematic storytelling, scene composition, shot sequencing, blocking, and directing techniques.
  • Exceptional ability to evaluate acting realism, micro-expressions, emotional beats, character interactions, eyelines, and continuity.
  • Strong understanding of screen direction, the 180-degree rule, spatial consistency, and visual continuity across multiple shots.
  • Hands-on familiarity with generative AI image and video tools such as Runway, Midjourney, Sora, Kling, or similar platforms.
  • Ability to independently create visual samples and reference scenarios for testing and comparing AI-generated outputs.
  • Strong written communication skills and ability to translate directing and performance observations into precise evaluation notes, captions, and structured feedback.
  • Ability to review and evaluate large volumes of AI-generated video with attention to creative and technical detail.
  • Strong analytical and observational skills combined with professional creative judgment.
  • Ability to collaborate with multidisciplinary teams of filmmakers, animators, researchers, and AI engineers.
  • Availability to work consistently during weekdays.

Nice To Haves

  • Access to a reliable 4K display for detailed, high-resolution visual review.

Responsibilities

  • Define creative quality standards for AI-generated video performances, including acting realism, movement, blocking, screen direction, eyelines, continuity, emotional nuance, and cinematic storytelling.
  • Review AI-generated video sequences as production dailies, identifying issues with acting, staging, character positioning, camera direction, and narrative flow.
  • Provide detailed, actionable feedback on performance quality, emotional authenticity, character interactions, visual storytelling, and directing choices.
  • Identify, generate, and curate reference scenes and challenging dramatic performances to test generative video models.
  • Assess cinematic continuity, including the 180-degree rule, eyeline matching, shot continuity, spatial relationships, and consistent character positioning.
  • Develop a structured taxonomy of common model failures, including unnatural acting, inconsistent blocking, incorrect screen direction, emotional mismatches, and continuity errors.
  • Help establish reviewer qualification criteria, assessment frameworks, and calibration guidelines.
  • Review captioning and annotation standards to ensure they capture directing intent, acting nuances, character interactions, staging, and emotional context.
  • Collaborate with AI researchers, engineers, and creative specialists to improve model performance, evaluation accuracy, and output quality.
  • Use generative image and video platforms to create reference samples, test scenarios, and comparative examples for model evaluation.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service