Full-Stack Dev to Build Low-Cost Image-to-HTML Tool (DeepSeek / MiMo)

We are looking for a developer to build a lightweight web application that takes AI-generated UI screenshots/mockups (e.g., from ChatGPT) and converts them into clean, responsive HTML and CSS (or Tailwind CSS). To keep operational and build costs low, the developer will use AI-assisted coding workflows (Cursor, Claude Code, etc) and integrate low-cost inference models (e.g., DeepSeek, Xiaomi MiMo, or hybrid vision pipelines) for the conversion backend. Key Responsibilities: Build a simple web dashboard with image drag-and-drop support. Integrate vision/LLM APIs tuned specifically to interpret AI-generated UI layouts, color palettes, and components. Create a side-by-side view: Original AI Image | Live Rendered HTML/CSS Preview | Copyable Code. Write clear system prompts to ensure output code uses modern Flexbox/Grid layouts and responsive structure. Requirements: - Hands-on experience integrating Vision / LLM APIs (OpenAI, DeepSeek, MiMo, or OpenRouter). - Strong background in React/Next.js and modern CSS / Tailwind CSS. - Uses AI coding tools (Cursor, AI agents, Claude Code) for fast, cost-efficient execution. To Apply: Briefly explain how you would structure the prompt/pipeline to accurately parse an AI-generated UI image into clean code.