Dexian Asia Pacific Hiring! Full Time AI Video Content Creator in - Ricebowl

We're looking for a creator who makes commercials that sell — and who wants to help build the tools that make them. About half your time is production: you take a product or campaign brief and deliver finished ads for paid social, CTV, short-form drama, gaming and brand campaigns — fast, and at the quality brands expect. The other half is making the product better. You'll label and evaluate AI-generated output, diagnose failure modes, and feed structured feedback to the engineers and designers sitting next to you. This isn't a side duty: you sit with the product R&D team, not in a separate content pod, and your taste becomes the quality bar the product ships against. You might come from years in advertising and production, or be an AI-native creator with a reel of ads that actually ran. Ideally both.

Commercial production — ~50%

Concept to final cut: briefs into scripts, storyboards and finished ads, from 6-second bumpers to 60-second spots — product-selling ads, brand films, short-form drama, gaming creatives and social content.

Translate the brief: turn brand objectives, audience insights, key messages and calls to action into clear creative executions.

Direct the models: precise prompts, reference images, control inputs and model settings to hit specific visual, motion, voice and brand outcomes. Add live-action or motion graphics where they help.

Cut for each platform: native versions for Reels, TikTok, YouTube Shorts, pre-roll and TV/CTV, with the right ratio, length, pacing and captions.

Build test variants: hooks, openings, CTAs, formats and markets, so performance teams can A/B test — then use the results on the next round.

Protect the brand: hold product accuracy, brand guidelines, legal, platform ad policy and AI-disclosure requirements in every deliverable.

Quality evaluation & labeling — ~30%

Label and evaluate output for visual quality, brand alignment, message clarity, audience appeal, voice naturalness, platform fit and effectiveness against the objective.

Name the failure modes: artifacts, inconsistent subjects or products, unnatural motion, weak lip sync, wrong text, off-brand results.

Compare models, settings and workflows on quality, speed, consistency and cost, and keep a quality log of recurring patterns.

Run and take part in blind A/B evaluations.

Product feedback & workflow design — ~20%

Document friction, bugs and bottlenecks as they happen, and propose features weighted against real production pain.

Help define what "good enough" means for AI-generated commercial content — and make it evaluable.

Design repeatable workflows and reusable skills across concept → generation → edit → audio → delivery, and document prompts and templates for other creators.

Must have

Commercial experience — our strongest signal. Ads made for real brands, products or clients under real budget and timeline constraints. We weigh this over technical novelty.

Both sides of the commercial equation: brand-building work and conversion work judged on CTR, CVR and ROAS — and a view on how you'd cut them differently.

Finished AI-assisted video in a portfolio or published work, covering the full workflow from brief through generation, editing, audio and export.

Judgment about models — which one to reach for on a given creative goal, and why.

Command of current generative video, image, voice, music and editing tools, and the habit of picking up new ones as they ship. The stack changes every few months; we care that you can evaluate and adopt a new tool quickly, not that you know today's list.

Fluency in how ads perform now: hook in 1–3 seconds, sound-off design, vertical-first, UGC and creator-led formats, clear product hero, obvious CTA.

Product sense: you can tell "this output is bad" from "the tool made this hard," say which one you hit, and suggest the fix. Feedback engineers can act on.

Visual and brand sensibility, plus the detail eye that catches wrong product details, inconsistent lighting or artifacts.

Awareness of AI-content risk: rights and likeness, music licensing, disclosure and labeling, product fidelity across generated shots.

Taste — you know when AI helps and when it cheapens the work.

Speed and organization on tight deadlines, across multiple variants and feedback rounds.