Contact person: Business Client

Phone:Show

Email:Show

Location: Thousand Oaks, United States

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I’m building Mall World—an AI-native, continually evolving “dream Internet” that grows from what people tell it about their night visions, memories, and emotional states while at Mall World. Whenever a user shares a dream journal entry, a childhood recollection, or even a quick note on how they’re feeling, the system should cluster those inputs, map their symbolism, and then let world-logic spin up new environments, entities, and a living canon that can shift or stabilise as the community keeps interacting. ie. a dream of a full parking lot with no spaces meaning they don't fit in. The state of the game should encourge and be positive.

Here’s the core I need you to architect and prototype:

• Data ingestion layer that accepts text and image snippets of user dreams, memories, and emotional states (including rich visual descriptions of the mall’s nooks and crannies).
• Clustering / similarity engine—think embeddings, vector databases, or any technique that helps detect patterns and affinities among submissions.
• Generative pipeline to create 3D or 2.5D environments, populate them with entities, and maintain story continuity. The emphasis is on environments, entities, and an evolving canon, not traditional “gameplay”.
• World-state management so the lore remains coherent even as new contributions pour in.
• Simple interactive front end where users can drop in their inputs and immediately see the world adjust.

I’m open on exact tools—PyTorch, TensorFlow, LangChain, Unity, Unreal, Godot, you name it—as long as you can justify the choice and move quickly toward a demonstrable prototype. Clean, well-commented code and a brief technical document outlining your approach will be the acceptance criteria.

If you thrive on experimental AI, emergent storytelling, and large-scale collaborative worlds, message me with a short outline of how you’d tackle the clustering logic and content generation flow, plus links to any similar projects you’ve shipped." (client-provided description)