Nine years building content systems for organizations that couldn't afford for them to fail. Before that, years on-site diagnosing industrial hydraulics and mail processing equipment across three continents. The throughline: if something isn't working the way it should, I want to understand it at the level where understanding actually matters; then build something better.
I was trained as a journalist, which means I was trained to believe that no story is self-explanatory. Someone has to do the work of finding the essential thing in a complex situation and making it legible to someone who wasn't in the room. That instinct doesn't leave you when the subject shifts to a router manual or an API endpoint; if anything, it becomes more important. The difference between documentation that users trust and documentation they abandon is usually a question of editorial judgment as much as technical accuracy.
Before the content operations work, I spent years as a field service technician; on-site, under pressure, diagnosing failures in industrial hydraulic equipment and high-speed mail processing systems for customers across North America, South America, the Middle East, and Southeast Asia. That kind of work teaches you to think in systems at the level where abstraction is a liability. You can't fake understanding how something actually works when it's broken and someone needs it fixed. That grounding has stayed with me through every role since: I approach a CCMS migration or a publication workflow problem the same way I approached a hydraulic circuit fault. Find where the system actually is, figure out why it's there, and build toward where it should be.
In my own time I go deep on ambitious attempts to break the mold. IBM's Workplace OS experiments in the 90s. The Honda NR750 and its bizarre oval piston engine. Sega's early swings at online gaming with the Saturn and Dreamcast. The 3D accelerator wars back when S3, Matrox, 3dfx, and PowerVR were duking it out with a nascent Nvidia. What draws me to all of it is consistent: someone tried to do something genuinely different, and the distance between what they attempted and what happened is worth understanding; often more so than the straightforward successes.
Whether it's a 20,000-topic knowledge base that AI systems can't reach, or a publication cycle that takes three weeks when it should take one day, or a hydraulic cylinder that's leaking for no obvious reason; I want to understand it at the source, find the gap between how it is and how it could be, and close that gap if I can.
The Problem: Documentation was siloed in a legacy on-premises CCMS that couldn't integrate with modern delivery systems or AI-assisted tools, creating a growing bottleneck as the product portfolio expanded.
The System: Led end-to-end migration of 20,000+ DITA XML topics from the legacy DITAToo platform to SaaS Heretto; building an API-driven, single-source-of-truth content layer designed for structured delivery to AI chatbots, voice bots, and support platforms.
The Result: Static documentation library transformed into a live input layer for AI-assisted customer support. The content architecture is designed to serve six downstream platforms without duplication or drift.
The Problem: A lack of high-fidelity data made it impossible to identify systemic bottlenecks, plan capacity accurately, or demonstrate the team's output to leadership.
The System: Architected a custom Jira-based project management ecosystem from the ground up to capture real-time team velocity, throughput, and capacity data across a globally distributed team.
The Result: Transformed a reactive operation into a data-driven one; establishing defensible baselines for planning, enabling resource requests that could be backed up with numbers, and making invisible work visible.
The Problem: Legacy publication cycles carried excessive cross-functional overhead and hard dependencies on other teams, taking 2–3 weeks to ship even minor content updates.
The System: Reengineered the end-to-end publication workflow to remove external dependencies and give the documentation team direct ownership of releases; a true continuous delivery model.
The Result: Publication cycle reduced from 3 weeks to a single day. Content ships at the speed of the product, not the speed of a cross-functional scheduling queue.
The Problem: Severe team attrition created dangerous single points of failure across critical documentation functions, placing product launch timelines and compliance requirements at risk.
The System: Cross-training frameworks to maintain coverage during the gap, then a strategic hiring effort; two writers positioned specifically to eliminate the single-contributor dependencies that had made the team fragile.
The Result: 100% operational continuity with zero launch delays throughout the attrition period. Team rebuilt to a more resilient foundation than before.
Leading a globally distributed team of writers and consultants through a period of significant transformation. The work spans people leadership, content strategy, and platform architecture; having led the CCMS migration that moved our documentation onto an API-driven content layer, the current focus is integrating that platform with AI chatbot and voice bot systems, consolidating a legacy knowledge repository into the new platform, and preparing to launch a unified customer-facing portal alongside a separate internal portal for support agents. The knowledge base this team maintains serves approximately two million unique visitors per month.
Specialized in DITA XML structured documentation and information architecture while directing contract writers and an SEO consultant prior to formal promotion. Authored and edited several hundred DITA XML topics contributing to a structured corpus of 20,000+ scoped for the current migration, and hand-authored knowledge base content directly in HTML within Salesforce Knowledge for four years. Served as an ongoing UX copy partner to product and engineering teams across consumer hardware and mobile software.
Documented the Modern Banking Platform, a componentized, API-first system built to replace monolithic core banking infrastructure; work that required hands-on testing of nearly everything, since inaccuracies in banking software carry real regulatory and financial consequences. Produced release notes running hundreds of pages across three major releases and numerous maintenance cycles, built topic-based help systems for a large and genuinely complex product surface, and edited REST API documentation for 100+ platform APIs on FIS's Code Connect marketplace. Embedded closely enough in Scaled Agile DevOps sprint cycles to catch documentation gaps before they reached client-facing release.
Came in as a Technical Service Representative, supporting field technicians and end users across North America, South America, the Middle East, and Southeast Asia; working from circuit diagrams, hydraulic schematics, and CAD files to diagnose problems in loading dock and restraint equipment remotely. Started contributing technical documentation in 2017, took full ownership of all customer-facing publications by 2018.
What followed was a systematic rebuild of a documentation library that had accumulated years of inconsistency: brought 80+ publications into ANSI Z535.6 safety compliance, led California Proposition 65 compliance work across technical and marketing materials, and initiated and managed the transition from bulk printed instruction manuals to digital QR code delivery; a project with cost, environmental, and customer experience benefits that also happened to be my first encounter with the question of how physical objects should talk to the people using them.
On-site technical service for Kern high-speed mail processing equipment across Wisconsin, Minnesota, and Oregon. Diagnosed hardware and software failures under production pressure, analyzed technical drawings and service documentation to identify faulty components, and consulted directly with Kern AG headquarters in Switzerland on complex escalations. The job was fundamentally about understanding how systems fail; explaining that understanding clearly to people who needed to trust the diagnosis. I didn't know it at the time, but that's still what I do.
The '92 Porsche 968 in my garage is a useful window into how I think. It's a technically accomplished car that arrived at the wrong moment, was priced below several less capable Porsches, and was consequently dismissed by most of the market. The gap between what it actually is and how it's perceived is, to me, genuinely interesting. I've spent significant time on a rotating cast of strange vehicles that made brave and often commercially unsuccessful choices. I've owned Mazda rotaries. I've owned Renaults almost nobody in the US has heard of. Ambitious attempts at doing things differently tend to teach you more than the straightforward successes, and I find them proportionally more interesting.
My music taste runs toward things that reward serious attention over casual familiarity. Sonic Youth, Polvo, Mission of Burma, and Lync for the ways they found to use rock instrumentation against itself. Fugazi for the principles as much as the music. Silver Jews and Smog for literary, devastatingly precise songwriting. Portishead, Acetone, Boards of Canada for the slower and more deliberate kind of weight. Anything that gets better the fifth time than it was the first.
I see a lot of films at the Roxie, the independent theater in the Mission District where I'm a member; David Lynch, Hal Hartley, Miranda July, anything where the form is being used to do something a different medium couldn't. Fill in the rest with good food, interesting bars, video games that reward thought over reflexes, and an enthusiasm for understanding how things work that I genuinely cannot seem to turn off. The Resident Evil series in particular; I went deep enough into an undocumented music bug in the PC HD Remaster that I wrote up the fix myself, and that guide has been live on Steam ever since. When something isn't explained and it should be, I end up explaining it.
Case studies documenting the architectural and operational work behind the projects above, alongside writing samples that show the craft underneath it.
How a 2–3 week cross-functional publication cycle got rebuilt from the ground up into a team-owned, single-day continuous delivery model; and what it took to get there.
PDF Case Study From Legacy to Live: Architecting a 20,000-Topic CCMS Migration at NETGEAREnd-to-end account of a migration from a 20-year-old on-premises CCMS to a modern SaaS platform; from internal advocacy through vendor evaluation, contract negotiation, and implementation.
Writing Sample Software Release NotesA representative example of enterprise release documentation, built to demonstrate structure and voice without disclosing NDA-protected work.
Writing Sample What Are WiFi Bands, and How Many Do I Need?A knowledge base concept article that explains a technical topic and reasons toward a recommendation based on the reader's actual needs.
Writing Sample Resident Evil HD Remaster: Fixed Background Music Mod GuideAn installation guide I wrote for an undocumented community fix, still live on Steam. When something isn't explained and it should be, I end up explaining it.
PDF ResumeFull career history, skills, and accomplishments. Updated 2026.