Turning AI strategy into everyday adoption.

Driving AI adoption at AARP

What I led. Communications, training and enablement workstreams inside a cross-functional enterprise program deploying AI-driven personalization across web, app and email — plus the AI assistants I built myself, the responsible-AI translation work, and the operating model that determines how the next platform adoption runs.

How I lead it. The through-line is change management, not the technology. I'm a Certified Problem and Change Manager, and the first questions I ask are the same whatever the capability is: who has to work differently, what will make them resist, and what keeps the new way in place after the launch energy fades. Answering those is what turns a rollout into adoption.

Rebuilding HR operations for surgery centers

The problem. Surgery centers were vetting clinical hires by hand — reading resumes and license PDFs, then checking each credential against state boards and exclusion lists one at a time. Slow, error-prone, and a real compliance exposure if a lapsed or sanctioned license slipped through.

What I built. One pipeline carries every applicant from raw resume to a verified, ranked shortlist. A custom Skill reads each resume and license PDF and pulls the structured fields — RN/CST/CRNA license number, NPI, specialties, and years in the OR. Claude in Chrome checks each license against the state board and the OIG exclusion list in real time, flagging anything expired, mismatched, or sanctioned. A plugin writes the scored, verified candidate back into the ATS and drafts the interview invite — no re-keying, no dropped files.

The result. Hundreds of hours of manual work saved. Automating the read-verify-rank pipeline took hand credential checks off the HR team's plate — turning slow, error-prone vetting into a verified, ranked shortlist, so the team spends its time on hiring instead of paperwork.

Scaling an online art store with automation

The problem. A growing online art store ran entirely on manual effort — every order processed, fulfilled, and followed up by hand, and marketing happening only when there was time for it. That ceiling capped how much the shop could sell.

What I built. I automated the store end to end, orchestrated in n8n. A new order routes straight to Gelato for print-on-demand production and shipping, with automated customer updates along the way — no manual re-keying. On the marketing side, Claude drafts listing copy and creative, Mailchimp runs the email campaigns and repeat-customer flows, and new work auto-posts to Meta, Instagram, and Pinterest — so demand keeps coming in on its own.

The result. Sales rose 265% in just two months — with far less hands-on time per order.