ContentCraftAI

🚀 Inspiration

Modern content creation is time-consuming, repetitive, and often disconnected from a brand’s core voice. We wanted to build an AI-powered solution that doesn’t just generate content, but thinks strategically — like a real creative team. That led us to develop ContentCraft AI, a multi-agent system where specialized AI agents collaborate, debate, and produce high-quality, brand-aligned content.

💡 What it does

ContentCraft AI is an intelligent content creation platform that enables users to:

• Create campaigns and define creative briefs.
• Analyze brand DNA using reference material.
• Watch AI agents debate in a simulated “Creative War Room.”
• Generate multi-format content (blog posts, tweets, captions, etc.).
• Revise, translate, audit, and optimize content instantly.
• Manage content strategy with a 7-day AI-powered calendar.
• Gain Creative XP with every interaction in a gamified dashboard.

It also includes a secure Admin Dashboard with analytics, content audits, and AI-assisted moderation tools.

🛠 How we built it

We used the following stack:

• Frontend: Next.js (App Router), React, TypeScript
• Styling: Tailwind CSS, ShadCN UI
• State/Auth: NextAuth.js, JWT, React Context
• Backend: Node.js, MongoDB, API Routes
• AI Agents & Flows: Genkit, Google Gemini (via @genkit-ai/googleai)
• Deployment: Firebase App Hosting

Code modularity and AI agent logic was handled in the /ai/flows/ folder, using Genkit to orchestrate decision flows and content pipelines.

🧩 Challenges we ran into

• Integrating multi-agent conversations in real-time was tricky — especially managing context across steps.
• Handling Next.js App Router edge cases (e.g., useSearchParams() needing suspense boundaries).
• Keeping state sync between creative flows and user actions without race conditions.
• Implementing secure role-based dashboards (editor vs admin) with NextAuth.js was complex.

🏆 Accomplishments that we're proud of

• Developed and deployed a fully functional creator dashboard in just 4 days.
• Created realistic AI debates between creative personas using Genkit.
• Delivered multi-step content refinement tools (revise, audit, translate, optimize).
• Built a gamified user experience with XP and feedback loops that reward creativity.

📚 What we learned

• How to architect a large-scale multi-agent AI system using Genkit and Gemini.
• Deep insights into Next.js App Router quirks and advanced SSR/CSR handling.
• The importance of feedback loops for improving AI generation quality.
• How to build a complete content platform from idea to deployment in a few days.

🔮 What's next for ContentCraft AI

• Add real-time collaboration for teams (multiple editors per campaign).
• Train custom AI personas using past content to better reflect the brand.
• Integrate with external publishing APIs (e.g., LinkedIn, Medium, Instagram).
• Offer premium AI tools (like viral copy generators or SEO auto-tuners).
• Launch a public beta for early adopters and get real-world feedback.

Built With

• genkit-+-google-gemini-ai
• mongodb
• next.js-15
• nextauth.js
• react-18
• react-hook-form
• recharts
• shadcn-ui-(radix-ui)
• tailwind-css-3
• typescript-5
• zod