Selected project
HiPeople
AI hiring platform for screening, assessments, interviews, and reference checks.
Helped ship early AI features and unified three applications into one frontend system.
HiPeople promotional image with an AI screening dashboard and high-match score, image 1 of 4
Technologies & Tools
Tasks
The problem
HiPeople's frontend was split across separate repositories, even though the product covered screening, assessments, interviews, and reference checks. Shared work was harder to maintain across three applications. The team was also adding AI-supported hiring workflows to the product.
What I owned
I helped ship early AI features for reference-check question generation, assessment suggestions, and resume screening. I partnered across the stack on hiring workflows that joined Node.js pipelines and PostgreSQL-backed tooling with React interfaces, custom hooks, optimistic updates, and strict TypeScript.
I consolidated three applications into a Yarn workspaces monorepo and built a shared headless React component library documented with Storybook. That work took two months. I also hired and informally led three frontend engineers, owned frontend planning and standards, mentored teammates, and supported hiring.
Decisions and tradeoffs
The monorepo and shared library gave the applications one place for reusable frontend code. I used React.lazy, Suspense, memoization, and profiling to improve perceived performance on the Vite and React Router stack.
How I checked it
I established shared quality practices with React Testing Library, Playwright, and Jest. The team used them across applications to check components and end-to-end workflows.
Outcomes
Three applications moved into one monorepo and shared component library in two months. The AI work shipped in the hiring product, and the testing practices made checks more consistent across applications.