AI Creative Lead to own the generative stack that turns approved scripts and storyboards into finished, on-brand video and static ads. This person is two things at once: a creative with a deep understanding of direct-response marketing, and a hands-on operator who lives in the AI tools daily. You run the back half of our pipeline, from character-locked stills through AI video, lip-sync, and voiceover to editor-ready footage, and you treat every step as a creative call as much as a technical one: what serves the story, what sells, and what will read as a real ad in the feed.
The role reports into the Head of Creative Production and partners with our Creative Directors and Editor/Designers. You know what makes a DR ad work, the hook, the pacing, the proof, and the turn to the offer, and you know which model(s) to use and why. You can hold character, wardrobe, and continuity across a whole spot, debug a broken render instead of re-rolling, and get renders through touchy platform filters. You have shipped AI-generated video that ran as paid creative, and you can tell real-ad output from output that just reads as AI.
We are a human-led shop that uses AI to move faster without diluting craft. You protect the craft bar rather than lowering it to what the model makes easy, and you stay current with a stack that turns over frequently, treating every tool pick as a short-term decision.
RESPONSIBILITIES - Own generation end-to-end for paid social and native creative: character-locked stills, AI video, lip-sync, and clean generated voiceover, from approved script and storyboard to editor-ready footage.
- Lock base characters first, then build a still per beat that references them, holding character, wardrobe, prop, and continuity consistency across an entire spot.
- Pick the right model for each job (hero vs. volume, native-audio vs. img-to-video) and keep a live view of a fast-moving tool stack, pinning exact model versions at build time.
- Scout, test, and onboard new models and tools as the stack evolves, and retire the ones that fall behind, so the team is always working with the best available.
- Get renders through platform and category filters. Diagnose and auto-fix a broken render once, then escalate, and keep a running swipe of safe phrasings and prompt patterns that pass.
- Operate and lightly orchestrate the tooling (fan-out, scheduling, asset movement) so many jobs run in parallel without a person babysitting each one.
- Own the generative-side handoff into post: deliver organized, tagged, editor-ready clips into the DAM to the production team's standards.
- Feed the shared learnings file (format, copy, render, and legal notes per job) and read it back at the next job's start so the system compounds instead of repeating mistakes.
- Partner with the Creative Directors and copywriters to keep generated output on-voice, on-persona, and matched beat-for-beat to the script (AV sync).
- Bring a direct-response creative eye to every job: push each render toward what sells, the hook, the pacing, the proof, and the turn to the offer, not just what looks good.
- Operate inside compliance rails for our regulated categories: no better-drug or results claims, coverage/qualify framing, required disclosures, and honest routing. Hold generation to the same gate as the rest of the pipeline.
- Manage generative tooling spend against render volume, and give the Head of Creative Production a real monthly cost model as volume scales.
- Build reusable assets (prompt libraries, character references, templates, and workflow docs) that let editors and strategists self-serve the routine work.
REQUIREMENTS - 4+ years in creative production, motion, VFX, or content, with at least the last 1 to 2 years hands-on in generative AI video and image production for real deliverables.
- A real grounding in direct-response creative. You understand how a DR ad is built (hook, problem, mechanism, proof, offer, and CTA) and can judge whether a cut will convert, not only whether it looks polished.
- Demonstrated generative craft: a reel or portfolio of AI-generated or AI-assisted video and stills, with an honest account of what you made vs. what the model made, and (ideally) how it performed.
- Deep, current fluency with the generative stack: multi-subject-consistency image models, AI video (native-audio and img-to-video), lip-sync, and AI voiceover. You have opinions on trade-offs, not just tool names.
- Character and continuity discipline: you can hold a consistent character across many shots and beats, not just generate one good frame.
- Prompt and troubleshooting fluency: you can diagnose why a render broke or a character drifted and fix it deliberately, rather than re-rolling until something lands.
- Comfort building light automation/orchestration around the tools (e.g. n8n, Gumloop, or comparable) to fan out and move assets. You don't need to be an engineer, but you're not afraid of the glue.
- A finishing eye: you know when AI output reads as a real ad vs. a tech demo, and you hold the line on craft.
- Demonstrated speed and volume in a performance, DR, or high-output content environment. Comfortable shipping many cuts and variants a week across multiple categories.
- Comfort working inside compliance rails (regulated-category claims, platform ad policy, partner brand guardrails).
- Strong collaboration across Creative, Production, Post, and Buying, and clear written communication for briefs, prompt docs, and learnings.
- Self-starter with an entrepreneurial mindset, comfortable in a start-up style operating environment where the tools change under you.
STRONG PLUSES - Generative video that has run as paid social or DR creative with attributable performance, not only brand or experimental work.
- Concepting or copywriting experience on DR ads, so you can improve a script or angle, not just execute it.
- Direct experience generating for regulated DR categories and getting their renders through platform filters.
- Video editing or motion/VFX background (Premiere, After Effects, DaVinci) that lets you hand editors better material and speak their language.
- Experience building reusable AI workflows for creative production: prompt libraries, character-reference systems, Claude Projects or custom GPTs, or agentic tooling that compounds team output.
- Familiarity with DAM systems (Air, Frame.io, or comparable) and production project-management tooling.
- Comfort reading performance dashboards (Looker, Tableau, in-platform reporting) to close the loop between what rendered and what converted.
- A track record of picking up brand-new tools fast and forming a real point of view on them, because this stack turns over monthly.
#LI-REMOTE #LI-NM1
