How abhisheksahoo15/ai-resume-builder works
Project type and purpose
• This is a full-stack web application (“AI Resume Builder”) combining a Python API backend, a machine-learning model for ATS scoring, web scrapers for job discovery, and a browser-based frontend.
• From the README: users create resumes, check ATS scores, and discover jobs. The API backend is implemented in FastAPI (and mentions Django, though code structure centers on FastAPI in main.py).
File structure
• Top-level:
• .github/workflows/: CI/CD definitions (main_ai-resume-builder.yml, main_hirefireai.yml)
• main.py: FastAPI application entrypoint
• requirements.txt: Python dependencies
• ml_model/ats_model.pkl: pickled ML model for ATS scoring
• scrapers/: modules to scrape job boards (Google, Indeed, LinkedIn)
• utils/: helper modules (save_to_csv.py, save_to_json.py)
• static/ and templates/: frontend assets (HTML templates, CSS, JS)
• test/: unit/UI tests (test_main.py, test.html)
• docs/: design guidelines and onboarding docs
README for additional insights
• Tech stack: FastAPI, (Django), PostgreSQL, Python ML, GitHub Actions, Terraform, Azure Web App
• CI/CD pipeline: feature-branch → test suite → GitHub Actions deploy via Terraform to Azure
• DevSecOps: OWASP, static code analysis, security scans in pipeline, IaC
• Client browser (UI) – HTML/CSS/JS in templates/ + static/
• API Gateway / Web Server – FastAPI app defined in main.py
• Business-logic modules
– Resume creation service (template rendering, PDF generation)
– ATS scoring service (loads ml_model/ats_model.pkl)
– Job-finder service (invokes scrapers/*.py)
• Data layer – PostgreSQL database (for user profiles, resumes, job caches)
• Utilities – CSV/JSON exporters (utils/)
• External services
– Indeed, Google Jobs, LinkedIn APIs (via scrapers)
• CI/CD & Infrastructure
– GitHub Actions workflows (.github/workflows/)
– Terraform IaC (manages Azure Web App & PostgreSQL)
b. Relationships & data flows
1. User → HTTP requests ← Browser
2. FastAPI routes
• /create_resume → template renderer → returns generated resume
• /ats_score → loads ML model → returns score
• /job_find → calls scraper modules → aggregates job listings
3. API ↔ PostgreSQL (read/write user data, resumes, logs)
4. Scrapers → External job boards → returns job data → API
5. CI/CD: Developer push → GitHub Actions → run tests (in test/) → Terraform applies → deploy to Azure
c. Architectural patterns & principles
• Layered architecture: Presentation (templates/static), API/Controller (FastAPI), Service layer (resume, scoring, scraping), Data layer (DB)
• Separation of concerns: distinct modules for ML, scraping, persistence
• Infrastructure as Code: Terraform-managed deployment
• Automated pipelines: GitHub Actions for testing, security scanning, deployment
d. Key technologies & frameworks
• FastAPI (Python) as web framework
• PostgreSQL as RDBMS
• pickle-based model (ml_model/ats_model.pkl) for ML inference
• Python scrapers for third-party integrations
• GitHub Actions + Terraform for CI/CD & IaC
• Azure Web App as hosting platform
Detail
• Individual backend subcomponents (Resume Service, ML Inference, Scraper modules).
• Security scan step in pipeline.