AI Creative Editor & UGC Producer

WHAT THIS IS We run a portfolio of companies across B2B SaaS, AI and automation technology, e-commerce and DTC supplements, and travel tech. We're building one shared creative engine: a pipeline that turns research into platform-native, conversion-focused creative at volume — static ads, AI UGC, avatar-led video, short-form social. You own the middle of that pipeline. You take a brief and a research pack and turn them into finished, ready-to-ship assets. No waiting on a shoot, a studio, or a creator roster. You generate, direct, edit, and deliver. This is ongoing, long-term work — not a one-off project. Expect 30 hours per week to start, with room to grow. WHAT YOU'LL ACTUALLY DO - Generate finished assets: static ad creative, image-to-video, AI UGC talking-head content, avatar-led scripts, b-roll, product-in-scene, motion overlays. - Direct at the shot level — framing, lighting, camera movement, lens character — instead of accepting first-pass generations. - Hold character, product, and brand consistency across a campaign: same talent, same packaging, same look, across dozens of clips. - Kill the AI tells: morphing hands and text, plastic skin, uncanny gaze, physics breaks, lighting discontinuity, over-smooth motion. You know which are fixable in-prompt, which are fixable in post, and which mean regenerate. - Cut natively for each platform — TikTok and TikTok Shop, Reels, Shorts, Meta feed, YouTube. - Build modular creative: hooks, bodies, and CTAs designed to recombine, so one concept produces a testable matrix. - Build and document prompt libraries so results are reproducible by someone other than you. WHO WE'RE LOOKING FOR Required: - A portfolio of AI-generated or AI-assisted creative that has ACTUALLY RUN — paid social, organic short-form, or an e-commerce storefront. Experiments are fine as supporting work; we want to see shipped assets. - Real prompt engineering skill across at least two image models and two video models. You can explain WHY a prompt worked. - Professional editing in a real NLE — Premiere, Resolve, After Effects, or CapCut Pro. You can fix in post what the model got wrong. - Direct-response instincts: hooks, retention, pattern interrupts, angle testing. You know why the first two seconds decide it. - Range: you've produced for more than one type of business. - Volume without collapse. You have systems — templates, presets, batching, naming — that make pace sustainable. - Strong written and spoken English, and the self-management to run your own week across time zones. Nice to have: - Supplements, health, wellness, or beauty (claims-sensitive categories). - TikTok Shop or affiliate-style creative. - On-camera or traditional UGC experience. - Workflow automation and API-level tool chaining, or comfort working alongside AI coding agents. - Sound design, voice cloning, delivery direction. TOOLING We run a defined internal toolset and will walk shortlisted candidates through it. We're not asking you to arrive knowing our specific tools — we're asking you to be fluent across the categories and able to pick up a new one in a week. You should be working confidently today across current-generation image models, text-to-video and image-to-video, avatar and lip-sync platforms, AI voice, motion control, and a professional editing suite. Tell us what you use and why you chose it. We'll tell you what we use if we progress. HOW TO APPLY — READ THIS PART We get a lot of proposals. Here's how to not be filtered out. 1. Answer the screening questions properly. Generic proposals get archived unread. 2. SEND US A VIDEO OF YOURSELF. This is required. Applications without one will not be considered — no exceptions, however strong the portfolio. Two to three minutes, self-recorded, on camera, speaking to us directly. Introduce yourself, tell us why this role, and walk us through ONE piece from your portfolio: which models you used, your prompt approach, what failed first, what you fixed in post. Unedited is fine. We are not scoring production value. We want to hear you think. Upload it unlisted to Loom, YouTube, or Google Drive and paste the link in the screening question. 3. Include your portfolio. For each piece, one line: what was AI-generated, what was edited or shot, where it ran. Start your proposal with the word BANANA so we know you read this.