Kanban-style job search and networking tracker
Live production application replacing spreadsheet-driven job search workflows; designed to scale to thousands of users on free-tier infrastructure with ~30-50KB per-user footprint
CareerClose is a career networking and interview tracking platform I built to replace the spreadsheet-driven job search workflow that most candidates fall back on. The core idea is simple: give job seekers a modern, Jira-inspired kanban interface to manage applications, networking contacts, and interview schedules in one place — without the weight or cost of enterprise CRM tooling.
The product is organized around three connected surfaces. The job applications board moves cards through stages (Applied → First Round → In Progress → Offer → Closed) with drag-and-drop ordering, round counters for multi-stage interview loops, and support for multiple interview dates per application. The networking board mirrors that model for relationship building, tracking contacts through Reach Out → In Progress → Connected with interaction counts, last-contact timestamps, and relationship strength indicators. A unified calendar view then pulls everything together — interview dates, networking follow-ups, and custom reminders — so users can see their entire pipeline at a glance and navigate directly back to the source card from any event.
Authentication is handled through NextAuth v5, supporting both Google OAuth and email/password with verification flows and secure JWT-based sessions. I also added a guest mode late in development, which was an intentional choice: I wanted people to be able to try the product without committing to account creation, which is a surprisingly common friction point for tools in this space. Email notifications and reminders are wired through Resend, and Vercel Cron handles scheduled jobs like follow-up reminders.
Architecturally, CareerClose is designed to be lightweight and free-tier friendly. Each user's data footprint is intentionally small — text metadata only, roughly 30-50KB per user — which lets the app comfortably scale to several thousand users on Supabase's free tier before hitting real cost pressure. User data is fully isolated at the database level through Prisma, and the entire stack is deployed on Vercel with Supabase as the Postgres provider. Beyond the core boards, I built out a small analytics dashboard (applications by stage, interview conversion, activity timeline), global search across all boards, and a changelog, tutorials, and in-app help flow to make the product feel polished rather than like a weekend prototype.
The project is live at careerclose.com. It's been a useful exercise in shipping a real end-to-end product — from design doc through authentication, payments-adjacent infrastructure, and production deployment — and in making the hundreds of small UX decisions that separate a side project from something people might actually use.