I am a student focused on artificial intelligence, natural language processing, and efficient local AI systems.
I am building Akane, a local AI VTuber designed to feel expressive, independent, and consistent across conversations.
The project includes:
- Local LLM inference on constrained hardware
- Long-term memory and contextual recall
- Autonomous mood and internal state systems
- Discord and Visual Studio Code integrations
- Prompt architecture and personality design
- Inference profiling and performance optimization
Languages: Python, C++ Technologies: PyTorch, NumPy, llama.cpp, Git
Areas of Interest: LLMs, NLP, transformers, local inference, AI agents, and distributed systems
- Transformer and attention architectures
- LLM inference optimization
- Memory and retrieval systems
- AI system architecture
- Secure connections between private AI systems and external information sources
A locally hosted conversational AI system combining personality, memory, autonomous internal state, and developer-tool integrations while remaining efficient enough to run on resource-constrained hardware.