🧠 Social.ai : Social Media Analytics Platform

📋 Table of Contents

1. Project Overview
2. System Architecture
3. Deployment
4. Frontend Implementation
5. Backend Implementation
6. LangFlow Integration

---

🚀 Project Overview

🎯 Objective

The goal of this project is to build an analytics module that analyzes engagement data from mock social media accounts using Langflow and DataStax integration.

🔑 Core Components

• DataStax Astra DB: Database for efficient data storage and retrieval
• Langflow: Workflow creation and GPT integration
• React.js: Frontend framework
• Node.js: Proxy backend
• Google Gemini: AI-powered insights generation

⭐ Key Features

• Real-time social media analytics
• GPT-powered insights generation
• AI-powered chatbot
• Custom metric tracking
• Post performance analysis
• Engagement metrics calculation

---

🏗️ System Architecture

The system is designed with a microservices-based architecture consisting of:

1. Frontend: Built using React.js for a dynamic and interactive user interface.
2. Backend: A Node.js proxy server that handles API requests and WebSocket connections.
3. Database: DataStax Astra DB for managing data storage and retrieval efficiently.
4. AI Integration: Langflow for creating workflows and integrating with Google Generative AI.

---

🌐 Deployment

📡 Live Application

• Production URL: https://super-rookie-assignment-social-ai.vercel.app
• 🎥 YouTube Video: https://www.youtube.com/watch?v=HvGaHkaEa80
• Platform: Vercel (Frontend) | Render (Backend)
• Status: ✅ Active

🛠️ Deployment Infrastructure

• Frontend: React.js hosted on Vercel
• Database: DataStax Astra DB
• AI Integration: Langflow and Google Gemini

---

🔧 Backend Implementation

The backend is built using Node.js and acts as a proxy server to manage API requests and WebSocket connections.

🔄 WebSocket Management

• Unique requestId assignment
• Connection tracking
• Real-time data streaming
• Error handling

📡 API Endpoints

1. Chat Endpoint
   • Handles client requests
   • Forwards to Langflow API
   • Streams responses
   • Error management

---

🤖 LangFlow Integration

LangFlow is used to create an end-to-end AI-powered flow that transforms raw data into actionable insights using:

• Google Generative AI
• NVIDIA Embeddings
• DataStax Astra DB

📋 Flow Breakdown

1️⃣ File Upload (MOCK_DATA.csv)

• The file upload serves as the primary data source.
• Any CSV file with relevant data can be used.

2️⃣ Split Text

• The uploaded file is divided into smaller chunks for easier processing.
  • Chunk Size: 300 characters
  • Chunk Overlap: 50 characters

3️⃣ Astra DB Integration

• The processed chunks are stored in Astra DB for efficient data retrieval.

4️⃣ NVIDIA Embeddings

• Text chunks are converted into vector representations using NVIDIA's Embedding Models.

5️⃣ Prompt Creation

• A structured prompt is created to guide the AI response.

6️⃣ Google Generative AI

• The Google Generative AI component generates insightful content.
  • Model used: gpt-1.5-pro

7️⃣ Chat Output

• The generated content is displayed in a chat format.

---