REASONING / ADAPTATION
AI systems
Exploring how intelligent systems can reason, adapt, and work together. From modular architectures to more efficient ways of learning.
01 / RESEARCH
Questions first.
Possibilities next.
Our research connects intelligent software with purposeful action and the physical world. We explore across disciplines, rather than treating them as separate destinations.
AREAS OF EXPLORATION
PLANNING / INTERACTION
Autonomous agents
Investigating systems that turn understanding into action. Planning, tool use, and collaboration across changing environments.
PERCEPTION / EMBODIMENT
Robotics
Studying the connection between intelligence and the physical world. Perception, embodied learning, and meaningful interaction.
APPLICATION / EXPERIENCE
AI apps
Exploring how AI becomes useful in everyday software. Building and studying applications, interfaces, and experiences that connect research with real needs.
SELECTED PUBLICATIONOPEN TO READ
MRT: Modular Reinforced Transformers — A Scalable Architecture for Ultra-Fast, Domain-Accurate LLM Systems
A modular architecture combining domain-specialist language models, routing, and adaptive reasoning. The paper describes the system design, training approach, and evaluation methodology.