CareerLift AI

Smarter Resumes. Stronger Futures

Comment

🧠 Inspiration

As a team, we noticed that students and early-career professionals often struggle to understand how well their résumé matches the jobs they’re applying for. Job postings list long skill requirements, but it’s unclear which ones truly matter or what the candidate should learn next. We wanted a tool that provides an objective score, highlights missing skills, and recommends real-world opportunities—all powered by AI. This inspired us to create CareerLift AI, a personalized résumé insights platform built with React, Firebase, and Gemini.

🎯 What it does

CareerLift AI analyzes a user’s résumé and target role to deliver:

  • A score out of 100 measuring job alignment
  • The top missing skills for the desired position
  • Recommended certifications, courses, and real-world opportunities
  • A clean, real-time dashboard powered by Firestore The app uses a strict JSON schema so the AI always returns structured, predictable results.

UN SDGs + Data for Social Good

Our project uses AI and real job-market data to advance social good and support key United Nations Sustainable Development Goals. By using Gemini to analyze résumés and BigQuery labor datasets to identify 2025 in-demand skills, we provide transparent, unbiased, and accessible career guidance for students and underserved communities. This directly supports SDG 4 (Quality Education) by offering personalized learning and certification pathways, SDG 8 (Decent Work and Economic Growth) by connecting people to job-ready skills, and SDG 10 (Reduced Inequalities) by reducing bias and ensuring equal access to career insights. By transforming public data into actionable career recommendations, our app demonstrates how responsible AI and open data can expand opportunity, empower youth, and create a more fair, data-driven future for all.

🛠️ How we built it

  • Developed the frontend using React, functional components, and lucide-react icons
  • Implemented Firebase Authentication (anonymous + token-based)
  • Stored all analyses per user in Firestore under a structured collection path
  • Used onSnapshot() to stream real-time dashboard updates
  • Powered the AI analysis through Gemini 2.5 Flash, using a custom JSON schema (ANALYSIS_SCHEMA) to ensure consistent scoring and recommendations
  • Built the system to gracefully handle missing configs, authentication timing issues, and async UI states

⚠️ Challenges we ran into

  • Missing or invalid Firebase config required graceful fallback logic
  • Authentication timing issues caused premature UI rendering, solved with an isAuthReady flag
  • Duplicate Firestore listeners during hot reloads, fixed with proper cleanup
  • Unpredictable AI formatting, resolved by enforcing a strict JSON schema
  • Loading and async state management, requiring guards, null checks, and fallback UI

🏆 Accomplishments that we’re proud of

  • Successfully integrated React, Firebase, and Gemini into a cohesive full-stack AI app
  • Built a fully functioning, real-time résumé insights dashboard
  • Created a reliable structured schema that stabilizes AI output
  • Designed a clean, intuitive UI using professional components
  • Made the app resilient even when backend configs fail
  • Developed a tool we would personally use for career preparation

📚 What we learned

  • How to build full-stack apps using React + Firebase
  • How to design and enforce reliable structured outputs with LLMs
  • Best practices for async state management and authentication flow
  • How to collaborate across roles for frontend, backend, and AI logic
  • Understanding how résumé content and job descriptions can be transformed into measurable skill vectors
  • The importance of actionable, structured feedback rather than generic AI responses

🚀 What’s next for CareerLift AI

  • Adding multi-résumé comparison to track improvement over time
  • Integrating LinkedIn job scraping to automatically match resumes to real postings
  • Visualizing skill gaps using charts and graphs
  • Adding a career roadmap generator that creates a personalized 30-60-90 day upskilling plan
  • Allowing users to save analyses and export them as PDF reports
  • Adding support for portfolio, GitHub, and project analysis
  • Expanding the AI to track progress and provide weekly improvement suggestions

Built With

Share this project:

Updates