Project Description
We need to generate a high volume (likely several thousand) of AI images to support both:
Our platform’s AI backend (training it to understand search filter combinations)
Consistent front-end imagery
Scope:
You will be provided with structured inputs across two categories:
Structure (e.g. Apartment, Semi-Detached, Mansion, Bungalow, etc.)
Style (e.g. 1970s, derelict, rustic, modern, etc.)
Your role will be to generate images that accurately represent combinations of these filters
(e.g. Apartment > 1970s, Semi-Detached > Derelict, etc.)
Requirements:
- Ability to generate high-quality, consistent AI imagery at scale
- Strong prompt-writing skills (we are currently using AI tools such as Higgsfield, but open to your preferred workflow)
- Maintain visual consistency across categories while ensuring outputs still feel contextually accurate
- Organise and label images clearly based on filter combinations
Workflow:
- Initial batch for review
- Feedback from our team
- Iteration and refinement where needed
- Full rollout once direction is approved
Please note there may also be a task where you then upload these into our backend, but this has not been approved yet.