DeepSlide AI System Enhances Presentation Delivery Beyond Slide Design
DeepSlide is a human-in-the-loop multi-agent system that optimizes the entire presentation process, including narrative planning, slide-script generation, and rehearsal support, rather than just creating visually plausible slides. It integrates a controllable logical-chain planner with per-node time budgets, a lightweight content-tree retriever for evidence grounding, Markov-style sequential rendering with style inheritance, and sandboxed execution with minimal repair. A dual-scoreboard benchmark separates static artifact quality from dynamic delivery excellence. The system was tested across 20 domains.
Key facts
- DeepSlide is a human-in-the-loop multi-agent system for presentations.
- It optimizes delivery process including pacing, narrative, and rehearsal.
- Integrates a controllable logical-chain planner with time budgets.
- Uses a lightweight content-tree retriever for evidence grounding.
- Employs Markov-style sequential rendering with style inheritance.
- Includes sandboxed execution with minimal repair for renderability.
- Introduces a dual-scoreboard benchmark for artifact and delivery quality.
- Tested across 20 domains.
Entities
Institutions
- arXiv