Skip to content

About

AI-powered presentation builder (before AI could do it)

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Latest commit

 

History

351 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 

Repository files navigation

SlidesPlus logo

SlidesPlus

AI-powered presentation builder — built before alternatives like Claude, ChatGPT, or other AI tools could generate full presentations on their own.

What it is

SlidesPlus builds complete presentations from an idea, a prompt, or a reference document, using AI to generate and edit every slide. Unlike a traditional PowerPoint generator, each slide is stored as HTML (1920×1080 sections), which allows for free-form design, animations, real typography, and layouts a classic slide engine can't achieve.

How it works

  1. AI generation — a wizard asks questions to understand the context (topic, audience, tone) and builds the presentation section by section, matching against ~45 base templates and generating the HTML for each slide with the chosen model.
  2. Visual editing — once generated, each slide can be edited from a visual editor (drag, resize, styling) without touching code.
  3. Direct code editing — for full control, each slide's HTML can be edited directly in an embedded Monaco editor.

There's also chat-based iteration on an existing presentation, public read-only views for sharing a deck, and real-time collaboration (Yjs) for multiple people editing at once.

AI providers used

  • Gemini (Google)
  • ChatGPT / GPT (OpenAI)
  • NVIDIA NIM — Llama 4, Qwen, DeepSeek, through an OpenAI-compatible endpoint

The backend selects (or lets you select) the model depending on the step in the flow (wizard, slide generation, chat).

Authentication

Email/password (JWT + bcrypt), Google login, and GitHub login. There's also a Spotify integration (not for login — it's used to embed music/album art inside slides).

Stack

  • Frontend: React 19 + Vite + TypeScript + Tailwind, Zustand, react-router-dom, Monaco Editor, GSAP/Framer Motion/Three.js for visual effects, PPTX/PDF export, i18n (ES/EN).
  • Backend: Node.js + Express, Supabase + Postgres, Passport (OAuth), Multer, Playwright for e2e tests.

Structure

frontend/   React + Vite + Tailwind SPA
backend/    Node.js + Express API (Supabase, AI providers, OAuth)

Running the project

Frontend

cd frontend
npm install
cp .env.example .env.local   # fill in values
npm run dev

Backend

cd backend
npm install
cp .env.example .env         # fill in values
npm run dev

Notes

  • Each package keeps its own package.json, .env.example, and vercel.json.
  • Environment files (.env, .env.local) are git-ignored — never committed.

About

AI-powered presentation builder (before AI could do it)

Topics

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages