Now
Updated 5 September 2026. A short, current record of the work and questions that have my attention.
I joined AI Enablement in mid-June, continuing as a Senior Software Engineer. Our team supports roughly 40 engineers.
I own our multi-agent pull request review tool and contribute to our AI software development standard. Four teams are onboarded to the standard; adoption and feedback are still developing.
See the AI review work →Handover brings client assets, presentations, and shareable deliveries together, with view and download tracking.
I built the backend from a UI-only prototype using Convex and Cloudflare R2. It is soft-live and invite-only, with a couple of testers.
Read about Handover →A PHPUnit-to-Pest experiment changed about 600,000 lines in four days of background AI work. The full test suite passes.
Further validation and adoption are still pending. I am also examining the trade-off between context, cost, and accuracy in AI review.
Read the experiments and their limits →This is the time-sensitive part of the site. It records my current direction without turning unfinished work into a case study.
For selected projects and measured results, use the Work page. For dated tests and their limits, use Experiments.