🤖 Personal AI Assistant Backend

A scalable backend for a personal AI chat assistant, built with ASP.NET Core following Clean Architecture. Users can create and chat with customizable AI characters powered by Google Gemini, with secure authentication, OTP verification, and Redis-backed caching.

---

📖 Overview

Personal AI Assistant lets users create custom AI "characters" (personality, tone, storyline) and hold persistent conversations with them. The backend handles auth, character/conversation management, chat history, and integrates with Google's Generative AI API for responses.

---

📦 Key Features

• 🔐 Authentication — JWT-based auth with OTP email verification for registration/password reset
• 🤖 AI Character Management — create and configure custom AI personalities (ModelCharacter)
• 🗣️ Conversation System — create, manage, and persist chat sessions between users and AI characters
• 🧠 Rich AI Settings — customizable emotion, storyline, and behavior per character
• 📊 Chat History — persistent storage and retrieval of past conversations
• 🛡️ Role & User Management — role-based authorization
• ☁️ Media Uploads — image handling via Cloudinary
• ⚡ Performance — Redis caching and endpoint rate limiting

---

🧠 Architecture

Clean Architecture, 4 layers:

Project	Responsibility
PersonalAIAssistant.Domain	Core entities, repository interfaces, domain models
PersonalAIAssistant.Application	Use cases, application services, DTOs, interfaces
PersonalAIAssistant.Infrastructure	EF Core DbContext, external service integrations (Redis, Cloudinary, Gemini, SMTP)
PersonalAIAssistant.API	Controllers, DI setup, middleware, rate limiting
PersonalAIAssistant.Test	Unit tests

---

🚀 Tech Stack

• .NET 10 / ASP.NET Core Web API
• Entity Framework Core + SQL Server
• Redis (distributed caching)
• JWT Authentication
• Mapster (DTO mapping)
• Swagger / Swashbuckle
• Cloudinary (media storage)
• Google Generative AI (Gemini)

---

⚙️ Prerequisites

• .NET 10 SDK
• SQL Server (local or remote instance)
• Redis (local instance or hosted)
• A Google Generative AI (Gemini) API key
• An SMTP account for sending OTP emails (e.g. Gmail App Password)
• (Optional) Cloudinary account for image uploads

---

🛠️ Getting Started

1. Clone the repository

git clone https://github.com/TungSke/PersonalAIChatAssistant_BE.git
cd PersonalAIChatAssistant_BE

2. Configure appsettings.json

Open PersonalAIAssistant.API/appsettings.json and fill in your own values:

{
  "ConnectionStrings": {
    "DefaultConnection": "Server=YOUR_SERVER;Database=PersonalAIAssistant;Uid=YOUR_UID;Pwd=YOUR_PASSWORD;TrustServerCertificate=True;",
    "RedisConnection": "localhost:6379"
  },
  "Jwt": {
    "Issuer": "https://localhost:7277",
    "Audience": "http://localhost:3000",
    "Key": "YOUR_JWT_SECRET_KEY"
  },
  "EmailSettings": {
    "SmtpServer": "smtp.gmail.com",
    "Port": 587,
    "SenderEmail": "your-email@gmail.com",
    "SenderName": "YOUR_APP_NAME",
    "Password": "YOUR_APP_PASSWORD"
  },
  "CloudinarySettings": {
    "CloudName": "YOUR_CLOUD_NAME",
    "ApiKey": "YOUR_API_KEY",
    "ApiSecret": "YOUR_API_SECRET"
  },
  "GenerativeAI": {
    "AIAPIKey": "YOUR_GEMINI_API_KEY",
    "AIAPIKey2": "YOUR_GEMINI_API_KEY_2",
    "PrimaryModel": "gemini-3.0-flash",
    "FallbackModel": "gemini-2.5-flash"
  }
}

⚠️ Don't commit real secrets. For production, use environment variables, User Secrets, or a secret manager instead of editing appsettings.json directly.

3. Restore dependencies

dotnet restore

4. Apply database migrations

cd PersonalAIAssistant.Infrastructure
dotnet ef database update --startup-project ../PersonalAIAssistant.API

5. Run the API

cd ../PersonalAIAssistant.API
dotnet run

The API will be available at https://localhost:7277 (or the port shown in your console), with Swagger UI at /swagger.

---

🐳 Running with Docker

docker build -t personal-ai-assistant .
docker run -p 8080:8080 -p 8081:8081 personal-ai-assistant

Note: the current Dockerfile base images are pinned to .NET 8 while the project targets .NET 10 — update the mcr.microsoft.com/dotnet/aspnet and sdk tags to match before deploying.

---

🧪 Testing

dotnet test

---

📚 API Documentation

Interactive API docs are available via Swagger once the app is running:

https://localhost:7277/swagger

---

📄 License

Distributed under the terms of the LICENSE file in this repository.