Generate
Agents turn research context and market observations into testable investment hypotheses.
Company
Grace Investment Machine builds agentic intelligence to solve markets.
GIM builds self-learning agent systems that generate investment hypotheses, validate them against market evidence, and improve through continuous research feedback.
We bring together experience across quantitative investing, frontier AI research, and large-scale systems engineering.
Team backgrounds span Millennium and other leading global investment firms; frontier AI organizations including DeepMind, Meta, and Microsoft; and Harvard University, the University of Oxford, Tsinghua University, Peking University, the University of Hong Kong, and other leading universities.
The research system
Agents turn research context and market observations into testable investment hypotheses.
Each hypothesis becomes code and is evaluated against data, robustness tests, and market evidence.
Successful and failed experiments become research memory, guiding the next cycle of discovery.