Skip to content

README_en.md

Latest commit

 

History

History
176 lines (135 loc) · 5.94 KB

README_en.md

File metadata and controls

176 lines (135 loc) · 5.94 KB

AI Sales Roleplay

Overview

A roleplay system for improving sales skills using generative AI. Users can develop practical sales skills through voice-based conversations with emotionally expressive AI. This system is designed for junior sales representatives, helping them improve their sales skills through interactive simulations with AI.

Key Features

  • Voice Conversation with AI: Natural conversations powered by Amazon Bedrock
  • Real-time Emotional Feedback: Visualization of anger meter, trust level, and progress
  • Diverse Scenarios: Customizable sales scenes
  • Detailed Analysis Reports: Improvement suggestions and feedback after each session
  • Video Analysis During Conversation: Analyzes video during sessions to verify effective eye contact and gestures
  • Compliance Violation Check: Identifies statements that violate compliance rules
  • Reference Check: Verifies whether user statements are based on reference materials

Technology Stack

Frontend

  • React 19 + TypeScript
  • Material UI 7
  • Vite 6 (Build tool)
  • AWS Amplify v6 (Authentication)
  • React Context API (State management)

Backend

  • AWS CDK (Infrastructure as Code)
  • AWS Lambda (Python 3.13) + API Gateway
  • Amazon Bedrock (Claude 3.5 Haiku, Claude Sonnet 4.5)
  • Amazon Polly (Text-to-speech)
  • Amazon Bedrock Guardrails (Compliance check)
  • DynamoDB + RDS PostgreSQL
  • Amazon S3 (PDF materials, audio files)
  • Amazon Cognito (Authentication)

Architecture

Setup

Prerequisites

  • Docker
  • Node.js 22.x or higher
  • Python 3.12 or higher
  • Latest AWS CLI
  • Latest AWS CDK

Deployment

Quick Deployment with AWS CloudShell

You can easily deploy without any prerequisites using AWS CloudShell:

  1. Log in to AWS Console and click the CloudShell icon (terminal mark) at the top of the screen

  2. Run the following commands

# Clone the repository
git clone https://github.com/aws-samples/sample-ai-sales-roleplay.git
cd sample-ai-sales-roleplay

# Run deployment script
chmod +x bin.sh
./bin.sh
  1. Deployment Options (optional)
# Disable self-registration feature
./bin.sh --disable-self-register

# Use a different region
export AWS_DEFAULT_REGION=ap-northeast-1
./bin.sh

# Specify individual models
./bin.sh --conversation-model "global.anthropic.claude-sonnet-4-5-20250929-v1:0"

# Detailed customization
./bin.sh --cdk-json-override '{"context":{"default":{"allowedSignUpEmailDomains":["example.com"]}}}'
  1. After deployment completes, you can access the application from the displayed URL

Manual Installation

  1. Clone the repository
git clone https://github.com/aws-samples/sample-ai-sales-roleplay.git
cd sample-ai-sales-roleplay
  1. Install dependencies
# Frontend
cd frontend
npm install

# Backend
cd ../cdk
npm install
  1. Environment Setup Refer to AI Sales Roleplay Environment Setup (Japanese)

Documentation

Deployment & Configuration

Feature Specifications

API & Technical Specifications

Operations & Cost

Note: Most documentation is currently in Japanese. English translations are planned for future updates.

Project Structure

├── frontend/                    # React application
│   ├── src/
│   │   ├── components/         # UI components
│   │   ├── pages/              # Application pages
│   │   ├── services/           # API services, authentication, etc.
│   │   ├── hooks/              # Custom React hooks
│   │   ├── types/              # TypeScript type definitions
│   │   ├── utils/              # Utility functions
│   │   ├── i18n/               # Internationalization settings
│   │   └── config/             # Configuration files
│   └── docs/                   # Frontend-specific documentation
├── cdk/                        # AWS CDK infrastructure code
│   ├── lib/
│   │   ├── constructs/         # Reusable CDK constructs
│   │   │   ├── api/            # API Gateway related
│   │   │   ├── storage/        # S3, DynamoDB related
│   │   │   └── compute/        # Lambda related
│   │   └── stacks/             # Deployable stacks
│   ├── lambda/                 # Lambda function implementations
│   │   ├── bedrock/            # Amazon Bedrock integration
│   │   ├── scoring/            # Scoring engine
│   │   ├── textToSpeech/       # Text-to-speech synthesis
│   │   ├── scenarios/          # Scenario management
│   │   ├── sessions/           # Session management
│   │   ├── guardrails/         # Guardrails management
│   │   ├── rankings/           # Ranking features
│   │   └── videos/             # Video processing
│   └── data/                   # Initial data (scenarios, Guardrails config)
├── docs/                       # Project documentation
│   ├── api/                    # API specifications
│   ├── cost/                   # Cost estimation
│   ├── custom-resources/       # Custom resource guides
│   └── features/               # Feature specifications
└── .kiro/                      # Kiro AI configuration files

Security

See CONTRIBUTING for more information.

License

This library is licensed under the MIT-0 License. See the LICENSE file.