JobForge
AI-powered job search management platform built entirely with Claude Code across 16 sprints. Features streaming cover letter generation via the Anthropic API, AI resume tailoring with keyword gap analysis, interview prep with confidence tracking, a pipeline Kanban board, SerpAPI job search integration, and a full analytics dashboard.
The Problem
Job searching is chaotic: scattered tabs, generic cover letters, no way to track what matters. Most tools either track applications OR generate AI content, but nothing does both in a coherent loop.
Approach
Built the full job search lifecycle in one app across 16 sprints using Claude Code as the primary development environment. Every AI feature (cover letters, resume tailoring, interview prep, weekly report summaries) routes through a Vercel serverless proxy to the Anthropic API. The pipeline Kanban uses @dnd-kit for drag-and-drop with optimistic updates and Supabase RLS for data security.
Outcome
A fully deployed, production application. The project demonstrated that Claude Code can serve as the primary development environment for a full-stack production app, not just a code suggestion tool.
What I Learned
Building with Claude Code revealed how AI pair programming changes the pace of shipping. The constraint of 400-line file caps and session documentation forced better architecture than I would have written alone.
Available for demo upon request.