Featured work
Career HQ
A working career operations system built around evidence, clear state and human handoffs.
I want to build systems people actually use. Career HQ is a deployed personal workspace that I use to manage my own job search, built with AI-assisted development. The problem: job discovery, application documents, employer forms and recruiting conversations live in separate places. A successful workflow needs to preserve what happened across those boundaries, rather than assume that a generated document or a clicked button proves completion. The system connects employer job feeds, evidence-based fit assessment, tailored résumés and cover letters, application history and read-only Gmail synchronization. Company-level conversations, pipeline charts and a needs-attention queue make the next action visible. Email matching is limited to applications with submission evidence. Browser execution runs through a connected Codex session. Versioned handoffs retain form answers, attachment checks, explicit approval states and receipt evidence. Credentials, CAPTCHAs, unknown personal facts and new binding agreements remain human steps. The ledger supports recovery from interrupted forms and employer validation failures, and avoids repeating a recorded submission. I carry the work from workflow discovery and interface design through TypeScript implementation, APIs, database state and deployment. I use Codex and ChatGPT during development, review changes, run relevant checks and verify the deployed behavior. Building this workflow has exposed practical problems in mobile navigation, application handoffs and mailbox recovery, which I continue to improve from actual use. Career HQ demonstrates product systems engineering, API integration, state management and practical AI-assisted delivery. It is a personal deployed system; broader customer adoption is not claimed. This public case study describes the product without exposing its private application records, mailbox or admin workspace.