Research library
Active studies and technical notes on unsupported claims, workflow scaffolds, evaluator calibration, evidence provenance, intent graphs, and AI reliability.
Explore researchResources
A focused library of controlled examples, active research, technical notes, and public software related to evidence assurance and AI reliability.
Core library
These pages show the method, controlled outputs, and current research status before asking you to book a conversation.
Active studies and technical notes on unsupported claims, workflow scaffolds, evaluator calibration, evidence provenance, intent graphs, and AI reliability.
Explore researchHow a workflow evidence design sprint reconstructs a synthetic OOS investigation record chain.
Open walkthroughHow a regulatory-change path is mapped from published amendment to cutover.
Open walkthroughHow claims are audited against supplied evidence on a synthetic product package.
Open walkthroughA controlled engagement sequence for diagnosis, scope freeze, evidence intake, analysis, challenge review, release, and decision workshops.
Read the processSoftware, research infrastructure, report-system controls, and public project documentation related to evidence analysis and AI reliability.
View GitHubPublication discipline
Research pages identify whether work is an active study, working paper, preprint, or technical note. Examples identify whether they are synthetic, redacted, or permissioned. A useful artifact should state what it demonstrates and what it does not.
Start with a non-confidential description of the workflow, claim set, or research question.