Modern AI systems are increasingly composed of distributed agent networks—a meta-agent handles incoming requests, while specialized sub-agents handle specific tasks. In some scenarios, human agents are also part of the loop. From a user perspective, it is often unclear who is currently responding, what this agent knows, and how the interaction shifts when more agents (human or artificial) join in.

You will:

• Analyze existing interaction designs for distributed AI agent systems (including human-in-the-loop setups)
• Develop design concepts or prototypes that make agent roles and transitions transparent
• Explore interaction mechanisms for handing over a conversation from AI tohuman (or between agents)

You need:

• Good programming skills
• Interest in interactive systems, AI interfaces, and multi-agent interaction
• Motivation to work independently on a creative and technically challenging project