PitchPlease

PitchPlease takes a static PDF pitch deck and turns it into a narrated video presentation with talking avatars, slide animations, and professional voice synthesis. Upload a deck, wait for the pipeline, get a shareable video.

The pipeline

Every deck flows through five stages:

1. Upload & extract — PDF parsing pulls content and structure from each slide
2. Narration — AI generates a presenter script, then a talking avatar delivers it via Sievedata/VEED.io with lip-sync
3. Visual enhancement — fal.ai adds slide animations and transitions between sections
4. Voice synthesis — ElevenLabs produces multiple voice style options for the narration
5. Compilation — everything gets stitched into a distributable video format

Real-time progress tracking shows exactly where a deck is in the pipeline.

Key features

• Drag-and-drop interface — upload PDFs up to 10MB with a responsive web UI
• Talking avatars — lip-synced AI presenters narrate each slide
• Multiple voice options — choose from premium ElevenLabs voice styles
• OAuth authentication — sign in with GitHub or Google
• One-click deploy — ships to Vercel with minimal configuration

Stack

Next.js 15, React 19, TypeScript, Tailwind CSS, shadcn/ui and Framer Motion on the frontend. Vercel Postgres with Drizzle ORM, Supabase, and Vercel Blob on the backend. ElevenLabs, fal.ai, Sievedata, and OpenAI for the AI pipeline.

---