I work on evidence-first approaches to cyber assurance and governed AI-assisted delivery.

The common pattern is simple: define the claim, establish what evidence should support it, verify the parts that can be checked deterministically, and keep consequential decisions explicit.

Control Assurance Case Study
A synthetic GRC case study showing how supplied evidence can be organised against controls and checked for missing, stale, scope-mismatched or unsupported material before assurance review.

Governed AI Delivery Exemplar
A runnable Python exemplar showing bounded AI-assisted work, explicit authority, deterministic verification, integrity evidence and an explicit approval boundary.

Both projects demonstrate the same underlying principle: evidence should support a consequential claim before that claim is accepted.