Laboratory for Assured AI Applications Development
Translational AI research at the University of Notre Dame
About
LA3D is housed at the University of Notre Dame's Center for Research Computing. Our work sits at the intersection of trust and engineering, converting theoretical AI advances into practical applications that are trustworthy, comprehensible, accessible, and actionable.
Our initiative AI Success Factors: Engineering Trust in Deployments focuses on the engineering and infrastructure that makes AI safe and reproducible for real research workflows. We build in the open: code, knowledge bases, and evaluation artifacts live on GitHub so that the community can audit, reuse, and extend them.
Projects
Nuggets
Weekly research presentations and notes on AI engineering, trustworthy ML, and applied research workflows. Published with Quarto.
LA3D-LLM-Agents
Federation of LLM-driven research agents. Each member pairs an
llm-wiki with a published Agent Card; agents collaborate via
three communication modes (ask, message, post).
Wiki-Grounded Research Agents (WGRA)
A durable, LLM-authored, queryable memory layer for scientific projects. Typed frontmatter edges reify to RDF; a shipped library of SPARQL queries surfaces hubs, criticism cascades, and orphan pages; verification and discipline gates enforce provenance at commit time.
Browse all public repositories on our GitHub organization page.