PAPER – A Workspace Where AI Agents Work Together
What I Built
PAPER is a multi-agent AI workspace designed to change how people interact with AI.
Instead of working with a single AI assistant, PAPER allows users to create multiple AI agents and organize them into teams. These agents can collaborate, exchange information, and work together toward a shared objective.
The idea behind PAPER is simple: What if AI agents could work like a real team instead of functioning as isolated chatbots?
I interpreted the "Ghost in the Machine" theme as giving intelligence and individuality to software agents. Each agent becomes part of a larger system, where collaboration between agents is what makes the system powerful.
PAPER aims to make AI collaboration feel less like chatting with a machine and more like managing a team of intelligent workers.
Demo
[Add deployed application URL]
[Add demo video URL]
Code
https://github.com/yogesh201sits/paper
How I Built It
PAPER is built using a modern TypeScript-based stack.
Technology stack:
- Next.js & React – Application framework and UI.
- TypeScript – Type-safe development.
- Tailwind CSS & shadcn/ui – Interface and component system.
- Clerk – Authentication and user management.
- Drizzle ORM & Neon PostgreSQL – Database and persistence.
- OpenRouter – Model access and AI integration.
- Vercel AI SDK – AI application infrastructure.
- Trigger.dev – Background task orchestration.
- Daytona – Planned sandbox infrastructure for agent execution.
The architecture is designed around multiple agents rather than a single assistant. The goal is to give users a workspace where agents can be created, managed, and brought together to accomplish more complex tasks.
Why Does Open Innovation Matter?
Open innovation matters because the future of AI shouldn't be controlled by a single model provider or platform.
Building PAPER around flexible model access gives the architecture room to evolve as new models and capabilities become available.
It also makes experimentation easier. Different models can be evaluated for different agent responsibilities instead of forcing every task through one model.
Open ecosystems make it possible for developers to build on existing infrastructure, experiment with different approaches, and contribute improvements without starting from scratch.
My Agent Session
[Add DevRelay agent session link if available]
The Spirit of It
I started thinking about a different way of using AI.
Today, most AI applications revolve around one conversation: you give instructions, the AI responds, and the interaction ends.
But complex work rarely happens that way. People collaborate, divide responsibilities, exchange ideas, and build on each other's work.
Why shouldn't AI work the same way?
That thought became PAPER.
I want to explore a future where creating an AI team is as natural as starting a conversation.
Prize Categories
[Add applicable partner categories from the challenge rules]
Top comments (0)