Alexander van Rossum

Fractional CTO practice for organizations that have outgrown their engineering capacity. I design the governance, infrastructure, and agent systems that make AI dependable in daily operations, and cover architecture, Salesforce, vendor evaluation, and security review. I also migrate WordPress brochure sites to Astro, which removes the PHP and plugin attack surface. I build and ship open-source macOS tools (SideMark, Panoptisana) under the same governance-first method.

Replaced an unfinished, broken WordPress brochure site with a new Astro build and a full redesign. Moving off WordPress removed the PHP and plugin attack surface, an unnecessary risk for most brochure sites, and the static build scores near-perfect on Google PageSpeed Insights. Built in two days, not counting the client's content work. Shipped October 2026.

A Salesforce program spanning five integrated systems (central CRM, identity and SSO, data collection, program delivery, and web presence) had stalled for three years with no documentation. I learned the platform from zero, untangled the existing org, and delivered the architecture and migration, then set up ongoing governance.

Led technical operations for a digital agency for about seven years: AWS infrastructure, Kubernetes orchestration, Salesforce architecture, AI-driven automation, security, and client-facing technical strategy. When a system broke and nobody knew where to start, I diagnosed the root cause, designed the fix, built the tooling, and documented it so the team could run it afterward.

Years of AWS provisioning across staff turnover had left idle services, stale access keys belonging to former staff, and no documentation. I ran an AI-assisted audit that was read-only by default, approved each remediation by hand, and produced the account documentation as a byproduct.

A production Kubernetes cluster running client workloads lost its control plane. No remote access keys existed, the cluster was years past end-of-life with no infrastructure as code, and TLS certificates were weeks from expiring. I designed a new AWS EKS architecture, built the automation tooling, and migrated all 260 deployments solo in under three months while carrying my existing workload.