Growth Video/Image Content Creator (Singapore)

Job Description

• AI-Driven Growth Video/Image Content Production: Proficiently use mainstream AI video/image generation models and editing tools to independently handle creative ideation, asset generation, and post-production (editing, motion graphics, audio-visual compositing, etc.) for growth ad videos and marketing images. Produce ad creatives that balance aesthetic quality and conversion effectiveness to support the content needs of overseas growth business.

• Build an AI Asset Production System: Develop an automated visual asset production pipeline based on AI Agents, covering core scenarios such as ad video generation and editing, marketing image generation and editing, and website page design. Drive the stable implementation of AI-driven large-scale asset production in growth business.

• Abstract High-Conversion Strategies: Deeply deconstruct the visual language of high-converting ad videos, websites, and image assets — including narrative pacing, camera language, opening hooks, audio-visual synergy, as well as layout, color, composition, and visual hierarchy — and translate them into reusable prompt strategies and Agent workflows to continuously improve asset conversion efficiency.

• Asset Performance Evaluation System: Establish a quantitative evaluation system that covers both "content quality" and "ad delivery performance". Provide quality scores and optimization suggestions through multi-dimensional assessment, and take ownership of the final output quality of growth assets.

• Scaled Quality Assurance: Address issues in the generation pipeline such as style inconsistency, brand deviation, character and visual coherence, and localization adaptation. Design effective strategies to ensure the stability and usability of batch-produced content.

• Cross-Team Collaboration: Integrate resources from algorithms (prompt/model application), design (aesthetic control), and operations (feedback iteration) to drive efficient execution of growth projects.

Job Requirements

• Experience & Background: Bachelor’s degree or above; candidates with backgrounds in film/TV production, digital media art, animation, visual communication, or advertising design are preferred. 5+ years of experience in ad video design, internet visual design, or overseas growth asset design; 2+ years of hands-on experience in AI-driven design, automated asset production, or Agent products. Experience in feed/short-video ad creative design, or overseas growth design for tool/content products is preferred.

• AI Application Capability: Deep understanding of the capabilities and boundaries of mainstream image/video generation foundation models and Agent tools. Ability to flexibly leverage foundation models, Agents, and other tools to solve business problems. Proven experience in video/image model orchestration and automated workflow construction is preferred.

• Business Acumen: Familiarity with overseas growth ad creative workflows (in-feed video ads, TikTok/YouTube/Meta ad assets, landing page and campaign page visuals, etc.). Strong ability to deconstruct narrative pacing, camera language, audio-visual synergy, layout composition, color systems, visual hierarchy, and image-copy coordination. Excellent aesthetic perception and visual language understanding, with awareness of overseas localized aesthetic differences.

• Systematic Logic Capability: Strong product abstraction ability to translate complex growth requirements into clear technical solutions and executable roadmap planning.

• Data Sensitivity: Skilled at problem attribution through data analysis. Able to independently analyze underlying issues in production pipelines and conversion performance and propose solutions.

• Industry Insight: Continuously track technology and product advances in image/video generation, design Agents, and intelligent workflows. Keenly assess how technological boundaries can expand growth efficiency.

Verified from the employer's public Greenhouse job collection. The provider-supplied title, description, location and application URL were normalized without inferring unstated facts.