careercompass

Navigate your career journey with AI-powered job matching and analytics.

Comment

🧭 About the Project

CareerCompass is an AI-powered, dual-role career platform built to transform the way employers and job seekers connect in today’s fast-paced, skills-driven market. The idea was born out of the need to address inefficiencies in hiring—such as the application black hole, skills mismatch, and recruiter overload—using intelligent automation and data-driven matching.

🔍 What Inspired Us

During countless job searches and internship hunts, we noticed a recurring pattern: poorly matched opportunities, slow responses, and minimal feedback. We wanted to create a system that didn’t just list jobs but understood them—along with the applicants behind every resume. Thus, CareerCompass was built to bridge this gap using AI, analytics, and automation.

🏗️ How We Built It

We developed CareerCompass using:

  • Next.js 15 App Router with TypeScript
  • Styled using Tailwind CSS and Radix UI
  • Firebase Firestore for database and real-time updates
  • Firebase Auth for secure, multi-role authentication
  • Cloudinary for file storage and delivery
  • Google Genkit and Gemini AI models for intelligent workflows
  • Brevo for transactional and marketing emails
  • Deployed on Vercel for performance and CI/CD

Key AI flows were structured as modular Genkit pipelines:

  • findAndRankCandidates
  • parseResume
  • enhanceText
  • sendApplicationStatusEmail

These flows utilize semantic scoring and prompt engineering to enhance user experiences.

🧠 What We Learned

  • The power and flexibility of semantic AI matching using Gemini
  • How to structure a real-time, scalable multi-role web application
  • How to integrate AI-enhanced UX into traditional workflows

🧗 Challenges We Faced

  • Dual-role UX Design: Managing two distinct experiences (employer vs job seeker) without compromising UX required custom routing and state management.
  • AI Integration: Designing reliable, explainable AI flows while handling latency, cost, and edge cases was complex.
  • Resume Parsing: Supporting PDF, DOCX, and TXT formats with consistent output required advanced extraction logic and fallback AI flows.
  • Email Automation: Ensuring deliverability and integration with status triggers was technically challenging at scale.

🛠️ Built With

Category Technologies Used
Frontend Next.js 15.3.3, React 18, TypeScript, Tailwind CSS, Radix UI, Zod
Backend Next.js API Routes, Firebase Firestore, Firebase Auth, Cloudinary
Authentication Google OAuth, Email/Password (Firebase)
AI / ML Google Genkit, Gemini Pro, Gemini Flash, Google AI Studio
Email Automation Brevo SMTP and Brevo API
Dev & Deployment Vercel, GitHub, ESLint, TypeScript Compiler
UI Components Lucide, CMDK, React Hook Form, Date-fns
Security Firestore Rules, MFA-ready Auth, OAuth auditing, session token expiration

Built With

  • brevo-api
  • brevo-smtp
  • cloudinary
  • cloudinary-api
  • css
  • firebase
  • firebase-auth
  • firebase-firestore
  • firebase-sdk
  • firebase-storage
  • gemini-flash
  • gemini-pro
  • github
  • google-ai-studio
  • google-genkit
  • google-oauth
  • javascript
  • json
  • lucide
  • next.js-15.3.3
  • radix-ui
  • react-18
  • react-context-api
  • react-hook-form
  • swr
  • tailwind-css
  • typescript
  • vercel
  • zod
Share this project:

Updates