Agent Design & Frameworks
Building the foundations of intelligent agents: robust architectures, advanced information retrieval, and performance optimization.
- Information Retrieval for Agents
- LLM-based Agent Architectures
- Performance & Efficiency
The Human-Agent Ally Lab pioneers research that bridges humans and intelligent agents — augmenting how we think, supporting high-stakes decisions, and preparing society for a future where humans and agents work, decide, and live together.
HAA (はぁ〜): our name embodies the sigh of relief and satisfaction — "Haa~, I'm so glad this exists!" — the feeling when human-agent collaboration truly works.
Interconnected research areas shaping how humans and intelligent agents build, collaborate, and coexist.
Building the foundations of intelligent agents: robust architectures, advanced information retrieval, and performance optimization.
Exploring how agents augment human cognition in complex problem-solving, high-stakes decision-making, and collaborative work.
Investigating how integrating intelligent agents transforms social structures, economies, and the norms we live by.
Home of a SIGIR Early Career Researcher Award recipient and an ANLP Outstanding Paper Award (top 2%), with work published at ACL, EMNLP, SIGIR, and WWW.
Active collaborations across Japan, Taiwan, Germany, the UK, India, Korea, France, Thailand, and beyond — plus leadership of ACL SIG-FinTech and the NTCIR evaluation series.
We founded and run the FinNLP workshop series, shared tasks at NTCIR and SemEval, and tutorials at AACL, EMNLP, ECAI, SIGIR, and LREC.
We're looking for passionate PhD students, research interns, and postdoctoral fellows who share our vision of a harmonious human-agent society.