A full-stack AI career platform connecting job seekers with providers — featuring resume parsing, smart job matching, skill gap analysis, 5-level progressive AI interviews, and real-time messaging.
Features • Screenshots • Tech Stack • Setup • Architecture
Career Genie is a dual-sided AI career platform built with Streamlit and powered by Groq's LLaMA 3.3 70B model. It serves two types of users:
- Job Seekers — Upload resume, get AI skill extraction, match with jobs, analyze gaps, and practice interviews
- Job Providers — Post jobs with custom AI interview configurations, rank candidates, and message applicants
| Feature | Description |
|---|---|
| 📄 Resume Parsing | Upload PDF or DOCX — AI extracts skills, experience, education, and job titles |
| 🎯 Smart Job Matching | AI scores your resume against job descriptions with match %, matched skills, and gaps |
| 📊 Skill Gap Analysis | Visual heatmap + bubble chart showing exactly what you're missing and why it hurts your profile |
| 🗺️ Ordered Learning Roadmap | Personalized step-by-step skill learning plan with sub-topics, resources, and progress tracking |
| 🎤 5-Level AI Mock Interviews | Progressive interview system from Foundations → Expert with behavioral, technical, debugging, coding, and system design questions |
| 📈 Progress Dashboard | Charts showing match score trends, before/after improvement, interview performance, and where you lack |
| 💬 Direct Messaging | Chat with job providers about your applications |
| Feature | Description |
|---|---|
| 📢 Job Posting | Post jobs with AI-auto-extracted required skills |
| 🎤 Interview Builder | Configure behavioral, technical, and coding question counts, marks per type, difficulty, and passing threshold |
| 🏆 Candidate Rankings | AI-ranked candidate table by match score with filter by status |
| 📥 Resume Download | Download applicant resumes directly |
| 📊 Analytics Dashboard | Applications by job, status breakdown pie chart, top candidates |
| 💬 Candidate Messaging | Message candidates directly from their profile |
Each screen below shows a key feature of Career Genie in action.
The entry point of Career Genie. Shows the platform's 5 core features, sign-up options for both Job Seekers and Providers, and platform stats (2,400+ jobs, 94% match accuracy). Built with Streamlit's dark theme.
The seeker's analytics home. Includes a skill coverage radar chart across 6 categories, match score trend line over time, application status donut chart, interview performance bars, and a "Top 3 Things to Improve Right Now" section — all backed by live PostgreSQL data.
Upload a PDF or DOCX resume and Groq LLaMA 3.3 70B automatically extracts 24 skills grouped by category (Data & ML, Backend, DevOps, Soft Skills), detects years of experience, education, job titles, and generates a professional AI summary.
Browse all active job postings with real-time AI match scores (0–100%). Each card shows matched skills in green and missing skills in red, the company, location, and a one-click Apply button. Filter by job type or search by role.
Type any role or pick from AI-suggested roles (filtered by career level: Intern / Junior / Mid / Senior). The AI produces a skill heatmap, priority bubble map, and detailed cards for each missing skill — showing exactly why it hurts your profile, why it matters, and how to fix it, with 3 curated resources per skill.
A fully personalized upskilling plan broken into 8 sequential steps — each with timeframe, dependency ("learn after X"), sub-topics to study in order, and a progress tracker (▶ Start → 🔄 In Progress → ✅ Done). Progress is saved to the database so it persists across sessions.
Choose from 5 progressive difficulty levels before starting a mock interview. Each level card shows its question types (conceptual, behavioral, debugging 🐛, coding 💻, system design 🌐, leadership 👥), difficulty, and marks. AI suggests roles from your resume as clickable buttons.
An active Level 3 interview question of type Debugging — includes a real broken Python code snippet with the bug embedded. The candidate types their answer explaining the root cause, the fix, and why it happens. A hint is available on demand. 🌱 Stretch Skill badge marks questions from missing skills.
After completing all levels, a full report shows: overall grade, score per question bar chart, AI narrative assessment with specific strengths and weaknesses, and a level-by-level breakdown with strong areas and what to revisit before the next attempt.
The employer's analytics view. Shows KPIs (active jobs, total applicants, shortlisted), a bar chart of applications per job, status breakdown pie chart, top 5 candidates by match score, and a live recent activity feed showing new applications and interview completions.
Providers configure exactly how candidates are tested per job: set the number of Behavioral / Technical / Coding questions, marks per type, difficulty level (Easy / Medium / Hard), and passing threshold. The system auto-calculates total marks. Required skills are AI-extracted from the job description automatically.
All applicants ranked by AI match score. Each row shows the candidate's name, match % with a progress bar, top skills, current application status, and interview result. Providers can filter by status, set a minimum match score, download resumes, or message candidates directly from this page.
| Layer | Technology |
|---|---|
| Frontend | Streamlit 1.32 |
| AI / LLM | Groq API — LLaMA 3.3 70B Versatile |
| Database | PostgreSQL 16 |
| Resume Parsing | PyMuPDF (PDF), python-docx (DOCX) |
| Charts | Plotly |
| Auth | bcrypt password hashing |
| Environment | python-dotenv |
- Python 3.10+
- PostgreSQL installed and running
- Groq API key — get one free at console.groq.com
git clone https://github.com/YOUR_USERNAME/career-genie.git
cd career-geniepython -m venv venv
# Windows
venv\Scripts\activate
# Mac/Linux/WSL
source venv/bin/activatepip install -r requirements.txt-- In psql or pgAdmin:
CREATE DATABASE career_genie;cp .env.example .envEdit .env:
GROQ_API_KEY=your_groq_api_key_here
DB_HOST=localhost
DB_PORT=5432
DB_NAME=career_genie
DB_USER=postgres
DB_PASSWORD=your_postgres_passwordstreamlit run app.pyOpen http://localhost:8501 — the app auto-creates all database tables on first run.
career_genie/
├── app.py # Main entry point + page routing
├── requirements.txt
├── .env.example # Environment variable template
│
├── auth/
│ ├── login.py # Login with bcrypt verification
│ └── signup.py # Signup with role selection
│
├── seeker/
│ ├── dashboard.py # Analytics dashboard with charts
│ ├── resume_upload.py # PDF/DOCX upload + AI parsing
│ ├── job_match.py # Browse jobs + match score + apply
│ ├── skill_gap.py # Gap analysis, heatmap, roadmap
│ ├── interview_room.py # 5-level AI mock interview system
│ └── messages.py # Chat inbox
│
├── provider/
│ ├── dashboard.py # Analytics + top candidates
│ ├── post_job.py # Post jobs + interview builder
│ ├── candidates.py # View, rank, download, message
│ └── messages.py # Chat inbox
│
└── shared/
├── db.py # PostgreSQL schema + all queries
├── groq_client.py # All Groq/LLM AI functions
├── resume_parser.py # Text extraction from PDF/DOCX
└── analytics.py # All Plotly chart builders
users — Auth, roles (seeker/provider)
seeker_profiles — Resume text, extracted skills
provider_profiles — Company info
job_postings — JD, required skills, interview config
applications — Seeker-job links with match scores
interviews — Questions, answers, scores per application
messages — Bi-directional chat between users
skill_gap_history — Historical gap analyses for trend tracking
roadmap_progress — Per-skill learning progress tracking
All AI is powered by Groq's LLaMA 3.3 70B Versatile model via structured JSON prompting.
| Function | Input | Output |
|---|---|---|
extract_skills_from_resume |
Resume text | Skills, experience, education, job titles |
compute_match_score |
Seeker skills + job skills | Score %, matched, missing, verdict |
analyze_skill_gap |
Skills + target role | Gap analysis with why deducted, how to fix, ordered roadmap |
suggest_roles_from_resume |
Skills + level filter | 12 role suggestions by career level |
generate_leveled_questions |
Role + skills + level 1-5 | Interview questions by type and difficulty |
evaluate_answer |
Question + answer | Score + feedback + strengths + improvements |
generate_level_feedback |
Level results | Readiness for next level, strong/weak areas |
get_upskilling_recommendations |
Skills + target role | Ordered roadmap + topics + certs + projects |
| Level | Name | Question Types | Difficulty |
|---|---|---|---|
| 1 🌱 | Foundations | Conceptual, Behavioral, Technical | Easy |
| 2 🌿 | Practical | Technical, Scenario, Behavioral | Easy-Medium |
| 3 ⚙️ | Problem Solving | Debugging, Use Case, Coding | Medium |
| 4 🔥 | Advanced | Architecture, Hard Coding, Debugging | Medium-Hard |
| 5 💎 | Expert | System Design, Leadership, Expert Code | Hard |
| Variable | Description | Example |
|---|---|---|
GROQ_API_KEY |
Your Groq API key | gsk_... |
DB_HOST |
PostgreSQL host | localhost |
DB_PORT |
PostgreSQL port | 5432 |
DB_NAME |
Database name | career_genie |
DB_USER |
Database user | postgres |
DB_PASSWORD |
Database password | yourpassword |
streamlit==1.32.0 # UI framework
groq==0.9.0 # LLM API client
httpx==0.27.2 # HTTP client (pinned for groq compatibility)
psycopg2-binary==2.9.9 # PostgreSQL adapter
pymupdf==1.24.0 # PDF text extraction
python-docx==1.1.0 # DOCX text extraction
plotly==5.20.0 # Interactive charts
pandas==2.2.1 # Data manipulation
bcrypt==4.1.2 # Password hashing
python-dotenv==1.0.1 # Environment variables
- Fork the repository
- Create a feature branch:
git checkout -b feature/amazing-feature - Commit your changes:
git commit -m 'Add amazing feature' - Push to the branch:
git push origin feature/amazing-feature - Open a Pull Request
This project is licensed under the MIT License — see the LICENSE file for details.