Product/UI Designer — AI-Powered Creative Platform

Speel is an AI-powered video creation platform. Users upload a product image or URL, and the AI generates UGC-style video ads — unboxing sequences, talking-head reviews, product demos — without manual editing or production teams. Our primary users are content agencies producing ad creatives at scale, but we're building for any creator. Think of the Figma model: built for professional teams, loved by individuals. We're looking for a product/UI designer to own the interface end-to-end.

Our design philosophy

We want to be Apple, not Samsung. That means: fewer controls exposed, smarter defaults, opinionated choices that reduce decision fatigue, and a premium feel that signals quality before the user generates anything. The interface should feel like a tool built by someone who deeply understands the creator's workflow — not a wrapper around AI APIs. We care about: fast time-to-value (the user knows what to do within seconds), clarity over cleverness, reducing cognitive load without removing capability, and visual quality that builds trust.

What you'll design

• A mode-based creative workspace — A single-screen interface with a top navigation bar (mode tabs), a left sidebar (model selection, settings, input configuration), a center gallery grid (generated media), and a sticky bottom bar (AI copilot chat, status, and controls). One screen, many states.
• AI copilot interaction patterns — A chat-based copilot that orchestrates the workspace in real-time. You'll design how the AI communicates what it's doing, how it asks for user decisions (multiple-choice approval flows), and how multi-step generation workflows feel as they execute progressively.
• Sub-mode selection and discovery — A modal system for browsing curated presets across AI models — each with preview thumbnails, model names, and descriptions. This needs to scale as models and presets grow without becoming overwhelming.
• Dynamic, context-aware UI — Settings, input slots, gallery actions, and available tools change based on the active mode and model. The interface must feel stable and clear even as its contents shift dynamically.