RetailMind AI

An Enterprise-Grade, AI-Powered Retail & Supply Chain Optimization Platform

---

🌍 Live Production Deployment

RetailMind AI is fully deployed and running in production on AWS EC2.
👉 View Live Demo: http://13.232.48.137

(Note: If you are testing the live deployment, please use the public sign-up or the default credentials below)

---

🌟 Introduction

RetailMind AI is a state-of-the-art, production-grade enterprise platform designed to solve modern commerce challenges: inventory stockouts, inefficient replenishment, supply chain bottlenecks, and volatile sales demand.

By combining a robust .NET 8 Clean Architecture backend, a high-performance React 19 + TypeScript frontend dashboard, a native FastAPI Machine Learning microservice, and Firebase Authentication, RetailMind AI provides retail operators with predictive intelligence and real-time operations control in a secure, containerized topology running natively on AWS.

---

🚀 Key Platform Features

🛒 Inventory & Order Management Engine

• Atomic Stock Deduction: Secure transaction boundary guarantees inventory is atomically reconciled when orders are placed, eliminating race conditions and over-selling.
• Low-Stock Alerting: High-performance category tracking with active triggers to alert personnel when inventory drops below safety thresholds (IsLowStock).

🧠 Predictive Machine Learning Engine

• Demand Forecasting: Custom Random Forest regression pipeline serving native predictions in Python. Generates estimated future sales volume based on product SKUs, temporal seasonality, target price, and active promotion status.
• Logistics & Delivery SLA Estimations: Gradient Boosting regressor predicting transit durations (in minutes) and classifying Transit SLA risk levels (Met vs. At Risk).

🔒 Enterprise Security & Firebase Authentication

• Firebase Auth Integration: Modern, secure identity management using Google Firebase to handle user sign-ups, log-ins, and session persistence.
• JWT Access & Identity Mapping: Secure token verification via the ASP.NET Core backend to associate Firebase UUIDs with internal PostgreSQL employee/role records.
• Global Security Middleware: Injection of HSTS, CSP, XSS protection, and complete Nginx proxy hardening.

⚡ Performance & Reliability

• Cache-Aside Redis Layer: Integrates distributed caching to speed up high-traffic reads using standard cache invalidation triggers.
• Dockerized Orchestration: Six distinct containers seamlessly communicating over an isolated retailmind_prod_network, managed entirely by Docker Compose.
• Nginx API Gateway: Unified entry point dynamically routing frontend, ASP.NET API, and FastAPI traffic.

---

🏛️ Production Cloud Architecture (AWS)

RetailMind AI is deployed using a full microservices mesh on an AWS EC2 Ubuntu Instance, coordinated via Docker Compose and unified through an Nginx Reverse Proxy.

graph TD
    classDef client fill:#eef2f7,stroke:#64748b,stroke-width:2px,color:#0f172a,rx:6px,ry:6px;
    classDef proxy fill:#f0fdf4,stroke:#16a34a,stroke-width:2px,color:#14532d,rx:6px,ry:6px;
    classDef backend fill:#fef2f2,stroke:#dc2626,stroke-width:2px,color:#7f1d1d,rx:6px,ry:6px;
    classDef ml fill:#f0fdfa,stroke:#0d9488,stroke-width:2px,color:#115e59,rx:6px,ry:6px;
    classDef database fill:#eff6ff,stroke:#2563eb,stroke-width:2px,color:#1e3a8a,rx:6px,ry:6px;
    classDef cloud fill:#fffbeb,stroke:#d97706,stroke-width:2px,color:#78350f,rx:6px,ry:6px;

    Internet((🌍 Internet Traffic))
    Internet -->|HTTP :80| Nginx

    subgraph AWS EC2 Instance [AWS EC2 / Docker Host]
        Nginx["🛡️ Nginx Gateway"]:::proxy
        Client["💻 React 19 Frontend Container"]:::client
        DotnetAPI["⚡ .NET 8 Backend Container"]:::backend
        PythonAPI["🧠 FastAPI ML Container"]:::ml
        PostgreSQL[("🗄️ PostgreSQL 15 Container")]:::database
        Redis[("⚡ Redis Container")]:::database

        Nginx -->|"/ (Root)"| Client
        Nginx -->|"/api/*"| DotnetAPI
        Nginx -->|"/ai/*"| PythonAPI

        DotnetAPI -->|EF Core Core Logic| PostgreSQL
        DotnetAPI -->|Cache-Aside| Redis
        DotnetAPI -->|Internal Network| PythonAPI
    end

    FirebaseAuth["🔥 Firebase Auth"]:::cloud
    Client -.->|OAuth / Token Request| FirebaseAuth
    DotnetAPI -.->|Token Verification| FirebaseAuth
    PythonAPI -.->|Firestore SDK| FirebaseAuth

    style Internet fill:#fff,stroke:#000,stroke-width:2px,rx:50,ry:50

Loading

---

🛠️ Technology Stack

Architecture Layer	Technology
Frontend Web	React 19, TypeScript, Vite, TailwindCSS v4, Recharts, Framer Motion
Backend Core	.NET 8.0 (C# 12) Web API, Entity Framework Core, Serilog
Machine Learning	Python 3.11, FastAPI, Scikit-Learn, Pandas, NumPy, Joblib
Databases	PostgreSQL 15, Redis 7 Alpine
Identity / Auth	Google Firebase Authentication, Firebase Admin SDK
Cloud & DevOps	AWS EC2 (Ubuntu), Docker Compose, Nginx Reverse Proxy

---

🏁 Run the Platform Locally (Docker)

To test the entire production-grade orchestration locally on your machine, you can run the Docker Compose stack.

📋 Prerequisites

1. Docker Desktop
2. A valid Firebase Project (you will need to generate a .env.production file and a firebase-adminsdk.json key).

🐳 Spin up the Mesh

# Clone the repository
git clone https://github.com/Manvith-kumar16/RetailMind-AI.git
cd RetailMind-AI

# Create your production environment file (ensure keys are filled)
cp .env.example .env.production

# Spin up all containers in detached mode
docker compose --env-file .env.production -f docker-compose.production.yml up -d --build

Service Endpoints

Once Docker completes the build, Nginx will dynamically map everything to your localhost:

• Client Dashboard (Web SPA): http://localhost
• Backend API Swagger: http://localhost/api/swagger
• Python FastAPI Docs: http://localhost/ai/docs

---

🚀 Deploying to AWS EC2

What you need to do after every push (currently)

1. SSH into your EC2 instance:

ssh -i ~/Downloads/retailmind-key.pem ubuntu@13.232.48.137

2. Then run:

cd ~/RetailMind-AI
git pull origin main

docker compose \
  --env-file .env.production \
  -f docker-compose.production.yml \
  up -d --build

This updates the application on EC2 with the latest code.

---

🔒 Default System Testing Credentials

If you are running the system locally and the database migrations have run, the system automatically seeds a testing account:

• Email: admin@retailmind.ai
• Password: Admin@123456!

(Note: For the live AWS deployment, relying on real Firebase Auth, please register a new account on the login page)

---

👨‍💻 Contributing

1. Fork the repository.
2. Create your feature branch (git checkout -b feature/NewFeature).
3. Commit your changes (git commit -m 'Add NewFeature').
4. Push to the branch (git push origin feature/NewFeature).
5. Open a Pull Request.

---

📄 License

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