Inspiration
The job market is changing faster than students can keep up. Many don't know which skills employers want, send the same generic resume everywhere, and have no professional network to ask for advice. We built job_ready for Prompt 2: helping people build in-demand skills and meaningful networks in one place.
What it does
- Career profile: a short questionnaire (education, skills, experience, preferences) is turned by AI into a structured career profile.
- Base resume: generated from the profile and downloadable as a PDF.
- Opportunity matching: jobs ranked by fit, with a score and a reason, filterable by local / remote / abroad and sorted by deadline.
- Tailored resume: when applying, the resume is rewritten for that job description and exported as a new PDF, plus suggestions for skills and experience to add.
- Events: upcoming workshops and networking events, soonest first.
- Mentor coffee chats: students browse mentors and request a chat; AI drafts a personalized icebreaker. Professionals sign up as mentors and manage requests from their own dashboard.
- Mock interview (in progress): AI generates role-specific questions and grades answers with feedback.
How we built it
- Frontend: React 18 + Vite + React Router, tested end-to-end with Playwright.
- Backend: Java + Spring Boot REST API (
/api/v1), layered controller / service / mapper with MyBatis, JWT authentication. - Database: MySQL.
- AI: Google Gemini API behind one
AiClientinterface. 8 features use it: profile generation, opportunity matching, base resume, resume tailoring, improvement suggestions, mentor icebreakers, interview questions, and answer grading. A stub client lets us run tests offline. - AI disclosure: besides Gemini in the product, we used Claude (Claude Code) as a coding assistant for scaffolding, API docs, and debugging.
Challenges we ran into
- Gemini models were sometimes unavailable or overloaded, so we added a fallback model, retries on 429/503, and timeouts.
- Getting reliable structured JSON out of the LLM for resumes and profiles.
- Keeping frontend and backend in sync while the API changed quickly; we kept a versioned API doc and an OpenAPI file.
- Designing two different experiences for students and mentors in one app.
What we learned
How to design an LLM-backed product around clear interfaces, how to turn messy user answers into structured data, and how to coordinate a 4-person team across frontend and backend in a weekend.
What's next
Real opportunity data sources, event registration, mentor points and rewards, and finishing mo
