Retrieval-Augmented Persistent Memory for Believable Non-Player Characters: A Per-Agent Memory and Real-Time Dialogue System in a Game Environment

A retrieval-augmented persistent memory for believable non-player characters: a per-agent memory and real-time dialogue system in a game environment.

A retrieval-augmented persistent memory for believable non-player characters: a per-agent memory and real-time dialogue system in a game environment. As a reference project it shows how a system like this is put together in practice, which parts matter and how they are tested, so that a developer can learn from it and build on it.

It suits learners who want to work with language models and text or speech: preparing text data, using or fine-tuning a model, evaluating it and building an interface around it.

The main parts are real-time processing of incoming data, retrieval over an indexed document collection that grounds the language model's answers, cooperating agents with defined roles and tools, text or speech data preparation, and model training, fine-tuning or prompting. The reference build uses LangChain, FAISS, LLM APIs, Apache Kafka, and Python, with a clear project structure, setup instructions and sample data so that it can be run and extended.