🚀 Fullstack Engineer Challenge – AI Content Workflow

Welcome to the Fullstack Engineer Challenge! 🤖📝
In this challenge, you'll help the fictional company ACME GLOBAL MEDIA build a system to manage the content creation and review workflow for their international campaigns — powered by AI.

🎯 Context

ACME GLOBAL MEDIA produces ads, micro-sites, and marketing materials in multiple languages.
Traditionally, creating and translating this content is slow and error-prone. They want to experiment with LLMs to:

• Generate initial content drafts (headlines, product descriptions, etc.).
• Translate and localize content into multiple languages.
• Extract structured data (keywords, tone, sentiment).
• Keep a review workflow where humans can accept, edit, or reject AI suggestions.

Your task is to build a simple system to:

• Manage campaigns (each with multiple content pieces).
• Generate AI-powered drafts for a content piece using OpenAI or Anthropic.
• Provide translation/localization suggestions via AI.
• Track a review state (Draft → Suggested by AI → Reviewed → Approved/Rejected).
• Show updates to all users in real-time.

📌 Requirements

⚙️ Tech Stack

⚡ Must Include - Use the following technologies, aligned with our tech stack:

• Backend: You can use any stack you're comfortable with, but we recommend:
  • TypeScript + NestJS (Fastify/Koa also valid)
  • Python + FastAPI (Flask/Django also valid)
  • Go + Fiber (Gin/Echo also valid)
• API: REST and/or GraphQL (justify your choice if only one)
• Frontend: React (Next.js, Remix, or Vite)
• Database: PostgreSQL (primary), MongoDB (optional if needed)
• Containerization: Docker (required)
• AI Integrations: OpenAI and/or Anthropic SDKs (required)
• Bonus: LangChain, Kafka, Redis, ArgoCD, Kubernetes

📦 Deliverables

📥 Your submission must be a Pull Request that includes:

• A backend API that supports:
  • Creating a campaign and its content pieces.
  • Generating AI drafts (titles, descriptions, translations).
  • Updating the review state of content.
  • Querying campaigns with their content and review states.
• A frontend built with React to:
  • Display a campaign dashboard.
  • Trigger AI draft generation.
  • Provide UI to review/edit/approve/reject drafts.
  • Show updates in real-time.
• Docker setup to run the entire app locally.
• A README.md with:
  • Setup instructions.
  • Tech decisions and tradeoffs.
  • If applicable, reasoning for REST, GraphQL, or both.
• A docs/ folder for any diagrams, workflows, or extra notes.

📂 Suggested Folder Structure

/
├── .github/
│   ├── workflows/
│   └── PULL_REQUEST_TEMPLATE.md
├── docs/
├── backend/
│   ├── src/
│   ├── test/
│   └── Dockerfile
├── frontend/
│   ├── src/
│   ├── public/
│   └── Dockerfile
├── compose.yml
├── .env.example
├── README.md
├── .prettierrc.js
├── eslint.config.mjs
└── ...

🌟 Nice to Have

💡 Bonus Points For:

• Using LangChain to chain AI tasks (generate → translate → summarize).
• Supporting multi-model comparison (OpenAI vs Anthropic).
• Real-time features with WebSockets, GraphQL Subscriptions, or SSE.
• Automated testing & GitHub Actions CI pipeline.
• Unit/integration tests for API or AI-related logic.
• Using Redis/Kafka for async event messaging.
• Deploy manifests for Kubernetes or ArgoCD.

🧪 Submission Guidelines

1. Fork this repository.
2. Create a feature branch for your implementation.
3. Commit your changes with meaningful commit messages.
4. Open a Pull Request following the provided template.
5. Our team will review and provide feedback.

✅ Evaluation Criteria

🔍 What we'll be looking at:

• Ability to work across the stack (NestJS/FastAPI/Go + PostgreSQL + React).
• Integration of AI features in a clean, modular way.
• Clear data modeling and workflow management.
• Human-in-the-loop UX for reviewing AI content.
• Documentation of assumptions, tradeoffs, and AI design choices.
• Creativity in using AI to enhance the workflow.

💬 Final Notes

This challenge is designed to be flexible. Some tips:

• If you’re stronger in backend, focus there but add a simple UI.
• If you’re stronger in frontend, ensure your backend has clean APIs.
• Time-box your work — we want to see how you think and solve problems, not perfection.
• Surprise us with creative uses of AI! 🎉

🏁 Good luck and have fun building!