LandPPT: an AI-driven presentation platform that automates research, slide generation, and explanation video creation

What it solves

LandPPT is an AI-driven platform that automates the creation of professional presentation slides. It eliminates the manual effort of structuring content, finding relevant imagery, and designing layouts by converting documents or simple topics into complete, ready-to-present slide decks.

How it works

The platform uses a multi-stage workflow (Requirement Confirmation $\rightarrow$ Outline Generation $\rightarrow$ TODO Tracking $\rightarrow$ PPT Generation) powered by Large Language Models (LLMs). It integrates various AI providers (OpenAI, Claude, Gemini, Ollama) to generate text, and combines three image sources—local libraries, web searches (Pixabay, Unsplash), and AI image generators (DALL-E, SiliconFlow)—to automatically match visuals to content. For deeper content, it uses search engines like Tavily and SearXNG to perform real-time web research.

Who it’s for

It is designed for anyone needing to create high-quality presentations quickly, including students, educators, and professionals who want to automate the research, drafting, and visual design phases of slide creation.

Highlights

• End-to-End Automation: Converts a topic or document into a full PPT, supporting parallel generation for speed.
• Deep Research Integration: Uses dual search engines to fetch and summarize real-time web information for more accurate content.
• Multimodal Output: Generates not only slides but also accompanying speech scripts and 1080p explanation videos with AI-generated audio (Edge-TTS).
• Flexible Export: Supports multiple formats including PDF, HTML, standard PPTX, image-based PPTX, DOCX, and Markdown.
• Extensible Infrastructure: Provides a REST API and OpenAI-compatible endpoints for automation, with Docker support for easy deployment.