Feeling AI Narrative Demo
Structured multi-character storytelling demo built around trust, memory, and world-state reasoning.
Project Overview
During my internship at Feeling AI (Shanghai Yunwu Tech), I led the development of an AI-driven multi-character narrative demo designed to test how large language models could support structured interactive storytelling.
Narrative Systems Design
I architected a Western-themed worldbuilding loop and designed the relationship network between characters, defining how story logic, role dynamics, and world-state progression could be coordinated across multiple NPCs.
Core Contribution
A key part of the work was defining semantic tagging rules and data-structure standards for dialogue, trust, and memory systems. This enabled the team to test more reliable narrative reasoning, character continuity, and state-aware interaction within the demo.
Outcome
The project successfully validated the structured reasoning potential of LLMs in interactive narrative design, showing how semantic organization and system architecture can make multi-character AI storytelling more coherent, legible, and scalable.
Gallery
