Skip to content

blanklavender/pitch-slapped

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PitchSlapped - AI Shark Tank Pitch Simulator

Think you've got what it takes? Pitch your startup to a panel of ruthless AI investors and find out.

PitchSlapped drops you into a high-pressure Shark Tank-style pitch room where three AI investor personas — James, Vidya, and Layla — grill you on your startup idea in real time. After your pitch, you receive an AI-generated report scoring your performance across five key dimensions.

Built at HackDavis 2026.

Features

  • Live AI Investor Panel — Three distinct investor personas with unique voices and personalities powered by ElevenLabs Conversational AI
  • Real-Time Audio-Visual Sync — Character-level alignment synchronizes judge avatar highlights with audio playback at millisecond precision
  • AI Pitch Analysis — Post-pitch scoring across Clarity, Persuasion, Market Knowledge, Confidence, and Financials using Google Gemini
  • Immersive UI — Dark Shark Tank aesthetic with spotlight animations, concentric ring effects on active speakers, and color-coded live transcripts

Tech Stack

Layer Technology
Frontend React 19, Vite 8, React Router 7
Backend Express 5 (Node.js)
Voice AI ElevenLabs Conversational AI (WebSocket)
Analysis Google Gemini 2.0 Flash
Hosting Vultr Cloud Compute
Domain GoDaddy Registry

Architecture

User speaks into mic
       |
       v
ElevenLabs Conversational AI (WebSocket)
  - 3 investor personas with distinct voice IDs
  - Real-time streaming audio + alignment data
       |
       v
React Frontend (PitchRoom)
  - Parses speaker tags (<James>, <Vidya>, <Layla>)
  - Builds character-level alignment timeline
  - Syncs judge avatar highlights with audio playback
  - Captures full transcript
       |
       v
Express Backend (/analyze-pitch)
  - Sends transcript to Gemini 2.0 Flash
  - Returns structured scores + feedback
       |
       v
React Frontend (PitchReport)
  - Displays conic-gradient score chart
  - Shows verdict, strengths, improvements

Getting Started

Prerequisites

  • Node.js 18+
  • ElevenLabs API key + configured Conversational AI agent
  • Google Gemini API key

Setup

git clone https://github.com/blanklavender/pitch-slapped.git
cd pitch-slapped
npm install

Create a .env file in the project root:

ELEVENLABS_API_KEY=your_elevenlabs_api_key
ELEVENLABS_AGENT_ID=your_agent_id
GEMINI_API_KEY=your_gemini_api_key

Run

Start the backend and frontend:

# Terminal 1 - Backend
node server/index.js

# Terminal 2 - Frontend
npm run dev

Open http://localhost:5173 in your browser.

How It Works

  1. Prep Room — Read the rules and mentally prepare your pitch
  2. Pitch Room — Deliver your pitch live to three AI investors who interrupt with tough questions
  3. Pitch Report — Receive your AI-generated scorecard with detailed feedback

Scoring Dimensions

Category Weight What It Measures
Clarity 20 pts How well you articulated your idea
Persuasion 20 pts Your ability to convince investors
Market Knowledge 20 pts Understanding of your target market
Confidence 20 pts Poise and conviction in delivery
Financials 20 pts Grasp of business model and numbers

Team

Built with no sleep and lots of determination at HackDavis 2026.

License

MIT

About

Get grilled and get better with AI investors in the club!

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors