ResumeAdapter.com - Instant ATS Optimization for Tech Roles
The Problem:
72% of dev resumes get rejected by bots before human review (Source)
Juniors waste hours guessing keywords like "Kubernetes" vs "Docker Swarm"
The Build:
Stack: Bolt.new (React) + Firebase + GPT-4-Turbo
Shipped MVP in 72 hours using AI-generated code (38% of codebase via Cursor.sh)
Traction:
522 signups (83% juniors)
4.7% conversion to $9.99/mo tier
22% returning users (7-day retention)
Hard Lessons:
PDF parsing costs 60% of dev time (finally used PyMuPDF)
Free-tier users convert best after 3 analyses (now gatekeeping earlier)
Devs care more about keyword gaps than design ("Just tell me what to add!")
Ask HN:
How would you improve the technical implementation?
