CUBOPS AI-COMPANION

Conversational Universe of Behavioral and Emotion Optimization for Personality Simulation (CUBEOPS)

This repository contains the source code for building a SaaS AI-COMPANION Platform using Next.js 13, React, Tailwind CSS, TypeScript, Prisma, PostgreSQL, and Stripe.

---

Table of Contents

• Features
• Prerequisites
• Installation
  • Environment Variables
  • Setup Prisma
  • Start the App
• Backend
• Available Commands
• Contributing
• License
• Contact

---

Features

Tailwind Design

• Fully styled using Tailwind CSS.
• Integrated with Tailwind animations for a polished user interface.

Full Responsiveness

• Optimized for seamless display across devices, ensuring a great user experience.

Authentication

• Utilizes Clerk Authentication with support for Email and Google.

Form Validation

• Implements react-hook-form for intuitive client-side form validation and management.

Error Handling

• Includes react-toast for server-side error handling and user notifications.

Page Loading State

• Ensures a smooth user experience with optimized page loading indicators.

Subscription Management

• Fully integrated with Stripe to handle subscription payments, including:
  • Monthly subscriptions
  • Token-based top-ups
  • Time-based call plans

Route Handlers

• Implements POST, GET, and DELETE API endpoints using Next.js route handlers.

Server-Side Data Fetching

• Fetches data directly in server-side React components, eliminating the need for intermediate APIs.

Component Relationships

• Manages relationships between parent and child components efficiently.

Layout Reusability

• Features reusable layouts for consistent and modular application design.

Next.js 13 App Router

• Organizes the application using the Next.js 13 App Router and modern folder structure.

Image Uploads

• Integrated with Cloudinary for secure and efficient image hosting and management.

LLM Integration

• Utilizes OpenAI APIs for advanced AI-driven features, such as personality simulation.

---

Prerequisites

• Node.js: Ensure you have Node.js 18.x.x installed.
• PostgreSQL Database: Set up a cloud-hosted PostgreSQL database for application data.
• Stripe Account: Obtain necessary API keys for subscription management.
• Clerk Account: Configure Clerk for authentication.

---

Installation

1. Clone the Repository

git clone https://github.com/VALASALARAKESH/AI-COMPANION-V2.git
cd AI-COMPANION-V2

2. Install Packages

Install all dependencies using npm:

npm install

---

Environment Variables

Create a .env file in the root directory and add the following variables:

Variable Name	Description
TOKENS_TOPUP_PRICE_ID	Price ID for tokens top-up.
CALLTIME_TOPUP_5_PRICE_ID	Price ID for 5 minutes call time top-up.
CALLTIME_TOPUP_10_PRICE_ID	Price ID for 10 minutes call time top-up.
CALLTIME_TOPUP_30_PRICE_ID	Price ID for 30 minutes call time top-up.
UNLIMITED_SUB_PRICE_ID	Price ID for unlimited subscription.
PRO_SUB_PRICE_ID	Price ID for pro subscription.
NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY	Publishable key for Clerk authentication.
CLERK_SECRET_KEY	Secret key for Clerk authentication.
DATABASE_URL	Connection string for PostgreSQL database.
NEXT_PUBLIC_APP_URL	Base URL of the application.
NEXT_PUBLIC_CLOUDINARY_CLOUD_NAME	Cloud name for Cloudinary image uploads.
NEXT_PUBLIC_CLOUDINARY_UPLOAD_PRESET	Upload preset for Cloudinary.
STRIPE_API_KEY	API key for Stripe payment processing.
STRIPE_WEBHOOK_SECRET	Webhook secret for Stripe.
NEXT_PUBLIC_OPENAI_API_KEY	API key for OpenAI integration.

---

Setup Prisma

1. Generate Prisma Client:

   npx prisma generate

2. Push the database schema to your PostgreSQL database:

   npx prisma db push

3. Seed the database with initial data:

   node scripts/seed_phone_voices_bolna.js

---

Start the App

Run the development server:

npm run dev

The application will be available at http://localhost:3000.

---

Backend

Repository

The backend for this project can be found at: modal-ai-companion

Backend Features

• Character-based AI interactions with configurable personalities
• YAML-based character configuration
• Image generation support for character avatars
• Streaming responses
• Modal-based deployment architecture
• FastAPI backend integration

Prerequisites

• Python 3.10 or higher
• Poetry for dependency management
• Modal.com account
• Required API keys:
  • DeepInfra API key
  • FAL.ai API key
  • OpenAI API key (optional)
  • Google Cloud Storage credentials

Installation

1. Clone the backend repository:

git clone https://github.com/VALASALARAKESH/modal-ai-companion.git
cd modal-ai-companion

2. Install dependencies using Poetry:

poetry install

Configuration

Required environment variables:

• API_KEY: Authentication token (used to authenticate with app endpoint)
• MODAL_TOKEN_ID: Modal authentication token ID
• MODAL_TOKEN_SECRET: Modal authentication secret
• Deep Infra API key: Added to Modal secrets
• FAL.ai API key: Added to Modal secrets
• GCP credentials: Added to Modal secrets for image storage

Usage

Start the server:

modal serve src.app

To deploy:

modal deploy src.app

Backend Project Structure

• src/
  • app.py: Main FastAPI application
  • agent/: Modal agent implementation
  • handlers/: Image and LLM handling logic
  • models/: Data models and schemas
• test/
  • characters/: Character YAML files
  • client.py: Test client implementation
  • character_loader.py: YAML character loading utilities

Backend API Endpoints

• /init_agent: Initialize or update an agent
• /prompt: Send prompts to the agent
• /generate_avatar: Generate character avatars

---

Available Commands

Run these commands with npm run [command]:

Command	Description
dev	Starts a development instance of the app.

---

Contributing

We welcome contributions! If you'd like to contribute, please:

1. Fork this repository.
2. Create a new branch with your feature or fix.
3. Submit a pull request for review.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---

Contact

For any questions or feedback, please contact Rakesh Valasala:

• Email: [email protected]
• LinkedIn: linkedin.com/in/rakeshvalasala

---