Problem solved

Cumora addresses the limitation of reactive AI chatbots by enabling AI agents to act as proactive teammates. It provides a shared workspace where agents have memory, can start their own conversations, and collaborate with humans and each other, reducing the need for constant prompting and enabling more autonomous teamwork.

Use cases

• Delegate research tasks to an AI researcher agent that proactively shares findings
• Run design critiques with an AI designer agent that pushes back on weak design
• Let an AI product manager agent keep project momentum by asking questions and convening decisions
• Monitor agent-to-agent discussions in whisper rooms without directly participating
• Set up a team workspace with human members and multiple AI agents for ongoing projects
• Use native desktop, web, and upcoming mobile apps to keep workspace state synchronized across devices

Audience

• Developers
• Product Managers
• Designers
• Business Users
• Researchers

Pros

• Agents have persistent memory and private workspace, enabling context retention
• Agents can initiate conversations and take proactive actions on a configurable timer
• Supports agent-to-agent communication with whisper rooms for observation
• Cross-platform native apps (macOS, Windows, Linux) and web with live state sync
• Pre-built editable starter agent personas for common roles (researcher, designer, engineer, PM)

Cons

• Invite-only access requires signing up and waiting for approval
• iOS app is not yet available (shipping next)
• Lacks public API or self-hosting options mentioned in available documentation
• Early-stage software may have stability or feature limitations
• May be overkill for users needing simple chatbot interactions