Web teams, useful software, and dependable delivery.
I lead technical teams and build useful software. My work connects people, priorities and technical decisions, from deciding what matters to checking what ships.
Twenty-plus years at the intersection of engineering, design and the work people actually need to do.
I set priorities, coach engineers and build technical ownership. That means working through scope and dependencies with other teams, then staying close enough to delivery to make useful decisions.
My independent work keeps me close to implementation. Each project starts with something a person needs to do.
Read my resumeStrategy and delivery belong in the same conversation.
Website Manager · Sep 2020–present
Lead web engineering and CRO. Coach engineers, develop technical owners, and shape quarterly priorities with Marketing, Product and RevOps. Led Pardot-to-Marketo integration and Vue modernization, and introduced practical AI workflows.
Senior Software Engineer / Full Stack Developer
Feb 2019–Sep 2020
Refactored a content API with AWS Lambda, Node.js and GraphQL. Extended headless WordPress with React and Gatsby, and contributed to reusable web components and automated testing.
2002 onward
Websites, software, integrations and the operational details that make them dependable. More than two decades of building and learning.
A team needs clear priorities.
A product needs a reason to exist.
These projects show the problems I work on, the choices behind the implementation and what is ready to use. Prototypes and internal builds are labeled separately.
Balance platform modernization and growth work. I lead priorities, technical ownership and hands-on delivery at JumpCloud.
Website Manager · September 2020–present
Read the leadership case study ↗Connect team intake and delivery. I built TIbot with duplicate detection, approvals and guarded Slack and Jira updates.
Built and in use · Public overview
Read the automation case study ↗Connect booking, staff workflows and operations. I built PolyshOS across web, API, worker and native clients.
Web/API/worker production · Native internal testing
Read the PolyshOS case study ↗Design, engineering
and growth at one table.
A company website has to work for its visitors and the teams responsible for it.
At JumpCloud, I lead web engineering and conversion work as Website Manager: shaping priorities, coaching engineers, building technical ownership and staying close to implementation. The work spans Vue modernization, Pardot-to-Marketo integration, reusable components and the systems behind content and measurement.
Earlier at TrainingPeaks, I helped refactor a content API around AWS, Node.js and GraphQL, extend headless WordPress with React and Gatsby, and bring reusable components and automated tests into delivery.
Connect the workflow.
Keep decisions accountable.
Useful agents need more than a good prompt. They need a place in the work.
I build specialist workflows for intake, audits, planning, engineering, security and delivery. TIbot brings that approach into Slack and Jira, with webhooks, scheduled work, deduplication, human approval and guarded updates. Reusable specialist skills connect intake and planning to implementation, review and handoff.
I’m developing Agent Workflow Learning to turn execution reviews into tested changes to instructions. PageInsights addresses a separate problem: understanding where visitors encounter friction on a page.
Find the right thing.
Make the next step work.
The useful part happens between the question, the source and the next action.
Client search work for House of Destiny combines WordPress, OpenAI embeddings, Supabase vector retrieval and keyword matching. Grounded summaries, filters, caching and analytics connect a visitor’s question to material they can inspect.
A calendar scheduling system connects Google Calendar and iCloud CalDAV with timezone-aware availability and meeting creation. Locks, request identifiers, conflict checks and recovery turn an attractive booking interface into something dependable.
One place for a vehicle’s history, modifications, media and the story behind the build.
Build profiles connect modifications, dyno results, photography and QR sharing. The event side brings organizer workflows, judging and offline support into the same system.
One workflow.
Every platform’s reality.
A publishing tool that treats the awkward platform details as part of the product.
OAuth connections, ordered carousels, JPEG normalization, optional FFmpeg processing and chunked TikTok uploads. R2 media storage, diagnostic history and delivery states help make failures explainable.
The business behind
the appointment.
An operating system for bookings, calendars, waitlists and the people keeping a salon moving.
A multi-tenant platform with role-based permissions and operational workflows. The web, API and background worker share a data model, with a native SwiftUI companion in development.
Hold. Talk.
Get back to it.
Dictation designed to fit around the work, with local recognition and careful text cleanup.
A SwiftUI macOS and iOS app with a compact HUD, keyboard workflows and a personal dictionary. Apple Speech, optional Parakeet Core ML and WhisperKit provide recognition options.
Faster implementation makes choosing the right work more important.
I balance growth work with modernization, reliability and technical debt. The roadmap has to account for team capacity and the people responsible for delivery.
In product work, the same question becomes more concrete. Can someone find the right source, book an appointment or dictate without losing their place?
My delivery workflows connect discovery and planning to implementation, testing and release. Specialist roles get the context and scope they need, with explicit ownership when work runs in parallel.
Approvals and guarded updates keep consequential decisions with people.
I require completion evidence and independent review before work advances. A working prototype is useful evidence about an idea. Production readiness still needs testing, clear ownership and a way to recover.
I label prototypes and isolated tests according to what they establish. Each answers a narrower question than a production rollout.
I’m developing a way to record execution reviews and feedback, then compare proposed instruction changes with a baseline before adoption or rollback.
The aim is to improve the workflow with evidence. It does not train model weights.
Tell me about your team, what you need to build, and where the work is getting stuck.