AI Developer Needed – Photorealistic Holiday Lighting Mock-Up Generator

We are an established professional holiday lighting company looking for an experienced AI/image-generation developer to build a tool that can quickly create **photorealistic holiday lighting mock-ups from customer photographs**. Our sales team regularly receives photographs of homes, commercial buildings, HOA entrances, monument signs, landscaping, palm trees, shopping centers, and other properties. We need to show customers what specific holiday lighting designs will look like before installation. Currently, creating these mock-ups with general-purpose AI image tools requires too many corrections. We want to develop a specialized system that follows our instructions consistently and generates professional sales-ready mock-ups quickly. How We Want It to Work The basic workflow should be extremely simple: 1. Upload a customer's photograph. 2. Enter written instructions describing exactly where lighting or decorations should be installed. 3. AI generates a realistic finished mock-up. 4. Make simple written revisions when necessary. 5. Download/save the finished image for use in proposals and presentations. Example instruction: "Convert the photograph to realistic twilight. Add warm-white C9 bulbs at 15-inch spacing along the front roofline only. Wrap the trunks of the three royal palms with warm-white mini lights up to the bottom of the husk. Do not add lighting anywhere else." The resulting image should preserve the original building, landscaping, architecture, signage, windows, doors, etc., while adding only the requested lighting/decorations. Critical Requirement: Instruction Following This is extremely important. The AI must **not creatively add decorations or lighting that were not requested.** If we say "roofline only," it should not add lights to columns, windows, landscaping, ridges, trees, or other areas. The objective is not simply to create attractive Christmas images. The objective is to create **accurate visual representations of the lighting package we intend to sell and install.** Lighting & Decoration Types The system should eventually understand our standard products and installation styles, including: * C9 and C7 LED bulbs * Specific bulb spacing such as 12", 15", 18", 24", and 36" * Warm white, pure white, cool white, red, green and other colors * Mini-light wraps on palm trunks * Mini lights following individual palm fronds * Mini lights in trees, hedges and shrubs * Roofline, gutterline and architectural outlining * Monument and entrance signs * Columns and arches * Wreaths * Garland * Bows * Commercial-grade holiday décor * Residential and large commercial properties Lighting should look like **real commercially installed holiday lighting**, not generic AI-generated Christmas decorations. For example, a C9 bulb should have realistic proportions and professional, even spacing. Image Requirements Finished images should: * Be photorealistic * Preserve the customer's original property * Accurately follow written instructions * Have professional and consistent bulb spacing * Show realistic light output/glow * Maintain correct perspective and scale * Convert daytime photographs to a realistic twilight/dusk appearance when requested * Avoid changing architecture or landscaping unnecessarily * Avoid adding objects that were not requested * Generate quickly enough to be practical for a sales team We may ultimately provide reference images showing exactly how our different lighting products and installation methods should look. Speed Is Important This will be used as a production sales tool. We may need to create many customer mock-ups during our busy season, so we are looking for a workflow where a salesperson can upload a photograph, enter instructions, and receive a usable result quickly. More importantly, we want to minimize the number of correction cycles required to get an approved image. What We Are NOT Looking For We are not looking for someone who simply knows how to connect a website to ChatGPT or another image-generation API. We need someone with genuine experience in areas such as: * Generative AI image editing * Image-to-image generation * Inpainting / masking * Computer vision * Diffusion models * ControlNet or similar structural-control technologies * AI image APIs * Prompt interpretation * Maintaining architectural geometry during image editing * Building specialized AI workflows around reference images or training data Experience with tools/models such as FLUX, Stable Diffusion, ComfyUI, OpenAI image models, or similar technologies is highly relevant. We are open to the developer recommending the best technical approach. Long-Term Goal The first version can be relatively simple, but our long-term goal is to develop a specialized **Holiday Lighting Design AI** that understands our company's design standards. Eventually, we would like the system to recognize commands such as: "Warm-white C9 roofline" "Pure-white mini lights on the three center palms" "Light the trunks only" "Add lights to approximately 20 fronds" "Add a 48-inch commercial wreath with red bow" …and consistently understand what those products should look like based on our standards and reference images. Potential Future Features Depending on feasibility, future versions could include: * Saved company lighting standards * Standard lighting/decor product library * Reference-image library * Customer/project folders * Before-and-after comparison * Multiple design variations * Ability to revise an existing mock-up without recreating the entire image * User accounts for salespeople * Integration with our CRM/proposal system * Mobile/tablet interface These features are not necessarily required for the initial MVP. Who We Want We are looking for someone who can **design the technical solution**, not simply follow a detailed specification. We want you to evaluate the problem and tell us what architecture/models/workflow will produce the best combination of: **Accuracy + consistency + photorealism + speed + reasonable operating cost.** There is potential for significant ongoing development if the initial system works well. IMPORTANT – Please Answer These Questions Please do not submit a generic AI-generated proposal. In your response, please answer: 1. What AI image-generation/editing technology would you use for this project and why? 2. How would you prevent the AI from changing areas of the customer's photograph that were not requested? 3. How would you maintain accurate placement and spacing of C9 bulbs and mini lights? 4. Would you recommend using an existing image-generation API, an open-source model/workflow, custom training/fine-tuning, or a combination? 5. How would you teach the system our specific holiday lighting standards? 6. What would you build for the initial MVP? 7. Approximately how long would you expect an individual mock-up to take to generate? 8. Please provide examples of previous AI image-editing or image-to-image systems you have built. **Please include links, screenshots, or demonstrations of relevant previous work.** Generic chatbot, automation, or basic website-development experience alone is not sufficient for this project.