Hi Team,
I'm building a multi-tenant AI platform where organizations can create and manage their own AI Agents, each with access to Composio connectors (e.g., Gmail, Google Drive, Notion, etc.). The platform supports n organizations, each with n agents.
I’m evaluating the best architectural approach for integrating Composio:
1. Should I create a separate Composio project for each organization to ensure better isolation and data boundaries?
2. Or should I use a single Composio project for all organizations and handle multi-tenancy and access control within my app?
The application uses ReactJS + TypeScript on the frontend and follows a microservice-based ExpressJS backend.
Our priorities are tenant isolation, scalability, and ease of connector management across organizations.
Would appreciate any best practices or guidance on project structure, token management, and connector isolation at scale.
Thanks!