AI-powered military-to-civilian resume generator.

1. Open 000-docs/032-QA-TEST-ope-veteran-flow-test-script.md
2. Follow each step, mark Pass/Fail
3. If something fails → Create GitHub Issue

1. Go to GitHub → Issues → New Issue
2. Pick a template:
   • QA Bug - something broke
   • QA UX - works but confusing
   • QA Missing - feature not there
3. Fill in the blanks, attach screenshot if helpful

1. Check assigned issues
2. Create branch: fix/issue-NUMBER
3. Fix the bug
4. Create PR → tag Jeremy for review

generator/
├── 000-docs/              # Project documentation (6767 + NNN standards)
├── src/                   # Source code
│   ├── generators/        # Format-specific generators
│   ├── parsers/           # Data parsers
│   ├── templates/         # Template engine
│   └── utils/             # Utility functions
├── templates/             # Resume templates
│   ├── modern/            # Modern template
│   ├── classic/           # Classic template
│   └── minimal/           # Minimal template
├── data/                  # Resume data files
│   ├── examples/          # Example resumes
│   └── schemas/           # JSON schemas
├── output/                # Generated resumes
├── tests/                 # Test suite
├── config/                # Configuration files
└── scripts/               # Automation scripts

pip install -r requirements.txt

# Generate resume from data file
python src/main.py --data data/my-resume.json --template modern --format pdf

# Generate all formats
python src/main.py --data data/my-resume.json --all

# Interactive mode
python src/main.py --interactive

Edit config/config.yaml to customize:

• Default templates
• Output settings
• AI integration
• Styling preferences

# Run tests
pytest tests/

# Lint code
flake8 src/

# Format code
black src/

The AI pipeline processes military documents and generates civilian resumes:

1. Document Upload: Candidates upload military documents (DD-214, ERB/ORB, evaluations)
2. Text Extraction: Worker extracts text from PDF, TXT, DOCX files
3. Vertex AI Processing: Gemini 1.5 Flash generates:
   • CandidateProfile: Structured military career data
   • GeneratedResume: Civilian-friendly resume content
4. Firestore Storage: Results stored in candidateProfiles/ and resumes/ collections

After AI generation, resumes are exported to downloadable formats:

Storage Path Convention:

candidates/{candidateId}/exports/{timestamp}-resume.pdf
candidates/{candidateId}/exports/{timestamp}-resume.docx

Libraries Used:

API Endpoints:

Team notifications to #operation-hired Slack channel when:

• New candidate intake: When a candidate submits documents and starts processing
• Resume ready: When AI generation and PDF/DOCX exports are complete

De-duplication: Each notification type uses a Firestore timestamp field to prevent duplicates:

• candidates/{id}.firstSlackNotifiedAt - New candidate notification
• resumes/{id}.resumeSlackNotifiedAt - Resume ready notification

Slack Message Format (Block Kit):

🆕 New Candidate Intake
Name: John Smith
Email: john.smith@email.com
Branch/Rank/MOS: Army / E-6 / 11B
[View Candidate] button

🎖 Resume Ready
Name: John Smith
Branch/Rank/MOS: Army / E-6 / 11B
Downloads: PDF, DOCX
[View Resume] button

Deploys all services and verifies the complete pipeline works in production.

Deployed Services (Dev):

Applies Operation Hired branding to the frontend for a professional, veteran-forward experience.

Theme & Branding:

• Primary color: Gold #C59141 (matches operationhired.com)
• Dark header/footer: #1a1a1a
• Typography: Roboto + Roboto Slab fonts
• Layout: Branded nav header and footer with Operation Hired identity

Frontend Routes:

Candidate Resume URL Pattern: /candidate/{candidateId}

• Shareable link for operations team
• Redirects to /intake/{id}/complete
• Can be bookmarked by candidates

UI Components:

• Step indicator across all pages (3 steps)
• Drag-and-drop file upload
• Real-time status updates
• Download buttons for PDF/DOCX

Layout Component: frontend/src/components/Layout.tsx

• Navigation: Home, Resume Generator, Admin, Hire Talent, Government Services, Contact
• Footer: Company info, quick links, address
• Mobile-responsive hamburger menu

Theme Variables: frontend/src/index.css

--primary-gold: #C59141;
--primary-dark: #1a1a1a;
--success-green: #38a169;
--error-red: #e53e3e;
--info-blue: #3182ce;

Internal dashboard for Operation Hired team to view candidates and monitor pipeline.

Admin Routes:

Admin Data Sources (Read-Only):

Admin Features:

• Real-time candidate list with status filtering
• Status summary cards (total, ready, processing, errors)
• Candidate detail view with:
  • Candidate info (name, email, branch, rank, MOS)
  • System info (ID, timestamps)
  • Uploaded documents list
  • Resume preview (summary, skills, experience)
  • PDF/DOCX download buttons
  • Link to candidate-facing page

Security Note: Admin pages are publicly accessible in dev. Production deployment should add Firebase Auth.

# GCP
GCP_PROJECT_ID=resume-gen-intent-dev
VERTEX_LOCATION=us-central1
GEMINI_MODEL_NAME=gemini-1.5-flash

# Firebase Storage
FIREBASE_STORAGE_BUCKET=resume-gen-intent-dev.firebasestorage.app

# Frontend
VITE_WORKER_URL=https://resume-worker-dev-96171099570.us-central1.run.app

# Slack Notifications (Phase 2.1)
SLACK_OPERATION_HIRED_WEBHOOK_URL=https://hooks.slack.com/services/xxx/yyy/zzz

# Worker App URL (for Slack links)
APP_BASE_URL=https://resume-gen-intent-dev.web.app

Deploy Worker to Cloud Run:

cd services/worker

# Build and push image
gcloud builds submit \
  --tag us-central1-docker.pkg.dev/resume-gen-intent-dev/resume-generator/worker:latest \
  --project resume-gen-intent-dev \
  ../..

# Deploy to Cloud Run
gcloud run deploy resume-worker-dev \
  --image us-central1-docker.pkg.dev/resume-gen-intent-dev/resume-generator/worker:latest \
  --platform managed \
  --region us-central1 \
  --project resume-gen-intent-dev \
  --allow-unauthenticated \
  --ingress all \
  --memory 2Gi \
  --cpu 2 \
  --timeout 300 \
  --set-env-vars "GCP_PROJECT_ID=resume-gen-intent-dev,VERTEX_LOCATION=us-central1,GEMINI_MODEL_NAME=gemini-1.5-flash,FIREBASE_STORAGE_BUCKET=resume-gen-intent-dev.firebasestorage.app,APP_BASE_URL=https://resume-gen-intent-dev.web.app"

Deploy Frontend to Firebase Hosting:

cd frontend

# Build production bundle
npm run build

# Deploy to Firebase Hosting
firebase deploy --only hosting --project resume-gen-intent-dev

Deploy Firestore/Storage Rules:

firebase deploy --only firestore:rules,storage --project resume-gen-intent-dev

The worker service uses Puppeteer for PDF generation with system Chromium:

# services/worker/Dockerfile uses node:20-slim with Chromium
FROM node:20-slim AS production

RUN apt-get update && apt-get install -y \
    chromium \
    fonts-liberation \
    ... # (additional dependencies)
    --no-install-recommends

ENV PUPPETEER_SKIP_CHROMIUM_DOWNLOAD=true
ENV PUPPETEER_EXECUTABLE_PATH=/usr/bin/chromium

Resource Requirements:

• Memory: 2Gi (for Chromium headless)
• CPU: 2 (for PDF rendering)
• Timeout: 300s (AI + PDF generation)

# Worker health
curl https://resume-worker-dev-96171099570.us-central1.run.app/health

# Expected response:
# {"status":"healthy","service":"worker","timestamp":"...","version":"0.1.0"}

See 000-docs/ for complete documentation following the 6767 standard.

MIT