Research area

Trustworthy AI, Reliability & Security

Reliability, security, interpretability, and transparent evaluation for AI and engineered systems.

Current research questions

What we are asking.

01

How can we identify and explain an AI system's errors and limitations?

02

How can trust and recovery be evaluated without hiding failure cases?

Research into systems

Related projects

RACoN

An experimental coordination layer for discovering, evaluating, and assigning computational workloads across heterogeneous machines.

Explore the system ↗Planned

SASS 4

A sample product slot for reliability, security, observability, and recovery-oriented tools.

Explore the system ↗

1 publications

Publications & manuscripts

Sample manuscript · Demonstration record2026

Failure-Aware Intelligent Systems: Designing Recovery Experiments for Distributed AI Infrastructure

Member 2 · Member 7 / Metademic Working Paper Series — Sample

This sample record demonstrates a reliability-focused publication format. It frames failure recovery as an experimental question and illustrates how a paper page can connect trust boundaries, recovery policies, observability, and repeatable fault-injection scenarios.

Read full manuscript ↗

Related software

Public research software will be linked alongside its documentation and release information.

Related datasets

No public datasets are currently listed for this area.