[Feature]: Multi-Agent Voice Breakout Rooms for Active Collaborative Learning

### Problem or Motivation

Current interactions in OpenMAIC primarily focus on a "Main Room" experience or linear QA. While the text-based Roundtable is effective for information retrieval, it lacks the high-engagement, low-friction environment needed for active pedagogical strategies. Students often learn better through verbal argumentation and multi-perspective debates, which are currently limited by text-only, single-agent interactions.

### Proposed Solution

I propose a "Voice Breakout Room" module that transforms the learning experience from passive QA to active discussion:
- Multi-Agent Presence: Instead of one-on-one tutoring, allow the creation of rooms where a student interacts with multiple AI personas simultaneously.
- Real-Time Voice-First UX: Integration of a low-latency voice interface (leveraging STT/TTS providers) to allow students to debate and discuss with agents naturally.
- Dynamic Grouping: A feature to split a "Course Session" into several "Breakout Rooms," each with a specific discussion prompt or role-play scenario.

### Alternatives Considered

_No response_

### Area

Multi-agent interaction

### Additional Context

This feature would shift OpenMAIC from a course-generation tool to a meta-cognitive platform. It encourages "Transversal Skill Development" by forcing students to articulate their thoughts under pressure from diverse AI perspectives.

### Case studies and applications

- Social-Emotional Learning (SEL): Developing empathy and conflict resolution through role-play with diverse AI personas.
- Language Acquisition: Improving fluency and pragmatic competence via natural, multi-agent voice immersion.
- Critical Thinking & Debate: Engaging in dialectical discussions to refine argumentative logic.