SmartCV

AI-powered resume builder. Runs entirely in your browser — no account, no cloud storage, no data leaving your device.

---

Preview

---

Features

• AI optimization — paste a job description and get an ATS match score, keyword gap analysis, and per-section rewrites
• ATS checker — instant client-side audit of contact completeness, content quality, formatting, and keyword coverage — no AI key needed
• Cover letter generator — AI writes a tailored cover letter from your resume and job description, with tone selection
• Job versions — fork your resume for a specific role, track application status (draft → offer), and compare the tailored copy against the original
• Undo / redo — bounded revision history persisted in localStorage; coalesced so fast typing creates one snapshot
• 10 resume styles — Classic, Modern, Executive, Minimal, Creative, Elegant, Academic, Split, Timeline, and a fully configurable Custom style
• Live theme colour — change the accent colour and the preview updates instantly
• PDF import — upload an existing resume; enable AI-powered parsing in Settings for higher accuracy
• One-click PDF export — pixel-perfect A4 PDF powered by Puppeteer
• Autosave — changes are debounced and persisted to IndexedDB
• Fully private — all data stays in the browser; the only outbound traffic is your own AI API calls
• Bring your own AI — configure OpenAI, Claude, Gemini, Grok, DeepSeek, Qianwen, Kimi, Doubao, or Wenyanyixin

---

Getting Started

Prerequisites: Node.js 20+, .NET SDK 10+

# 1. Clone
git clone https://github.com/ttang1024/Smart_CV.git && cd Smart_CV

# 2. Start backend (http://localhost:5173)
cd SmartCV.API && dotnet run

# 3. Start frontend (http://localhost:5173) — in a second terminal
cd SmartCV.Web && npm install && npm run dev

Then go to Settings to add an API key for your preferred AI provider.

---

Production Build

The React app outputs directly to SmartCV.API/wwwroot/, so a single .NET process serves both the API and the SPA.

cd SmartCV.Web && npm run build
cd ../SmartCV.API && dotnet publish -c Release -o ./publish
./publish/SmartCV.API   # or SmartCV.API.exe on Windows

---

Tech Stack

Layer	Technology
Frontend	Next.js 15 (App Router, static export), React 19, TypeScript
Styling	Tailwind CSS
State	Zustand, IndexedDB (idb), localStorage
PDF export	PuppeteerSharp (headless Chromium)
PDF import	pdfjs-dist + Tesseract.js OCR + optional AI parsing
Backend	.NET 10 Minimal API
i18n	i18next (en / es / zh-CN / zh-TW)
Deployment	Docker → Azure Container Registry → Azure Container Apps

---

Architecture

Browser (Next.js SSG)
  │  IndexedDB      — resumes, optimization sessions
  │  localStorage   — AI settings, job applications, revision history
  └─ /api ──► .NET 10 Minimal API
                  ├─ /api/ai/chat      — proxies to AI provider (avoids CORS, hides keys)
                  ├─ /api/pdf/parse    — extracts sections from uploaded PDFs (PdfPig)
                  └─ /api/pdf/generate — renders HTML to A4 PDF (Puppeteer/Chromium)

See tech/ for detailed module documentation.

---

Deployment

The app runs as a Docker container on Azure Container Apps. The image is built locally with Docker, pushed to Azure Container Registry, and served through Container Apps ingress.

Deploy infrastructure + app (first time or after infra changes):

./deploy.sh          # deploys with tag "latest"
./deploy.sh v1.2.0   # deploys with a specific tag

deploy.sh does four things in order:

1. az group create — creates or reuses the smart-cv-rg resource group
2. az acr create — creates or reuses the private Azure Container Registry
3. docker build + docker push — builds the Docker image locally and pushes it to the registry
4. az deployment group create — provisions or updates the Container Apps environment and Container App from azure/main.bicep

Deploy app only (after frontend/backend code changes):

./deploy-app.sh          # builds and deploys with tag "latest"
./deploy-app.sh v1.2.0   # builds and deploys with a specific tag

deploy-app.sh skips the Bicep infrastructure step and just runs docker build, docker push, and az containerapp update. Use this for faster iteration when only application code has changed.

Prerequisites: Docker plus Azure CLI logged in (az login) with Contributor access to the resource group.

---

Contributing

Open an issue before submitting a large PR. For small fixes, a PR is fine directly.

git checkout -b feat/your-feature
cd SmartCV.Web && npm run lint
# open PR against main

---

License

MIT