Advancing AI that serves humanity through rigorous research and practical applications.
Exploring approaches to give AI systems better memory and context awareness for more effective interactions.
Reducing computational cost without sacrificing capability through intelligent architecture design.
Building AI that complements human cognition and enhances decision-making while respecting autonomy.
Frameworks for building AI that helps rather than harms — from privacy-preserving architectures to anti-manipulation patterns.
We conduct research at the intersection of AI, cognitive science, and practical engineering.
Our work is guided by a simple principle: good research leads to good products, and good products make research meaningful.
We believe AI research should advance both the science and the practical applications that help people in their daily lives.
A Systems Engineering Approach to Empirical Prompt Testing
We contribute to the broader AI research community through publications in leading conferences and journals.
Where appropriate, we share our research methodologies to advance the field and enable reproducible research.
We partner with academic institutions and industry leaders to advance AI research for everyone.
Standards and governance documents that guide how we design, execute, and publish our engineering research.
New to research papers? This plain-language guide explains the statistical methods, terminology, and conventions used in our engineering studies. Terms from the companion guide are highlighted with dotted underlines in our papers — hover to see definitions.
Interested in collaboration or want to learn more about our research?
Get in Touchresearch@engramforge.com