slide-generation

Emblem launches as an AI tool that generates entire PowerPoint decks from a data room, claiming 80 slides in 35 minutes with traceable citations, supporting multiple AI models.

演示如何通过一次API调用，使用Gemini Managed Agents结合远程MCP服务器和GitHub技能，构建一个自主的金融分析师，自动生成包含自定义SVG图表的幻灯片。

An analysis arguing that AI presentation tools should focus on post-generation controllability for precise local edits, rather than just improving first-draft quality.

This paper proposes Spire, a framework that formulates slide personalization as an inverse planning problem, using structural denoising and reinforcement learning to infer latent design intents without relying on explicit templates or verbose instructions.

X+Slides is a new benchmark for evaluating audience-conditioned slide generation from source documents, using source-grounded probes and audience-specific utility weights. Experiments on DeepPresenter, SlideTailor, and NotebookLM show that current systems recover substantial but incomplete audience-essential information.

This paper introduces MemSlides, a hierarchical memory framework for personalized slide generation that separates long-term user profiles, working memory for session constraints, and tool memory for localized edits, enabling multi-turn revision without full regeneration.

The user seeks an AI tool to convert messy notes, meeting transcripts, and documents into readable slides for presentations.

A developer shares how they reduced their AI agent's weekly cost from $200 to $40 by routing simple subtasks to cheaper models like DeepSeek V4 Pro and Tencent Hunyuan while keeping complex reasoning on Opus 4.7, achieving comparable output quality for most work.

Turn Codex into an agentic slide builder with open-slide, allowing users to prompt once and have an agent ship a finished deck without leaving Codex.

DeepSlide is a human-in-the-loop multi-agent system for the full presentation process, from requirement elicitation and time-budgeted narrative planning to evidence-grounded slide-script generation and rehearsal support. It introduces a dual-scoreboard benchmark separating static artifact quality from dynamic delivery excellence, and achieves gains in narrative flow, pacing precision, and slide-script synergy.

This paper examines using generative AI tools (NotebookLM, Claude, M365 Copilot, Cursor, Claude Code) to generate slides from instructor notes, finding that coding assistants produce the best slides and that students cannot reliably distinguish AI-generated slides from human-created ones.