InsightStream - Customer Feedback Analyzer

InsightStream is an AI-powered analytics dashboard that transforms raw customer feedback into actionable business intelligence using Google's Gemini API.

🚀 Features

• AI-Powered Analysis: Uses gemini-3-pro-preview to deeply understand customer sentiment and intent.
• Sentiment & Segmentation: Automatically classifies feedback as Positive, Neutral, or Negative and segments customers (Promoter, Passive, Detractor).
• Strategic Insights: Generates concrete action plans, including "Next 7 Days Priorities" and "Risk of Inaction".
• Batch Processing: Paste multiple lines of feedback to get aggregate insights and per-entry breakdowns.
• Visual Dashboard: Interactive charts and clean data visualization using Recharts.
• Documentation: Built-in guide for users within the app.

🛠️ Tech Stack

• Frontend: React 19, TypeScript
• Styling: Tailwind CSS
• AI: Google GenAI SDK (@google/genai)
• Visualization: Recharts
• Icons: Custom SVG icons

📦 Setup & Usage

1. API Key: This application requires a valid Google Gemini API Key.

   • The app expects process.env.API_KEY to be available in the execution environment.

2. Input Methods:

   • Paste Text: Copy-paste raw feedback. For batch analysis, separate entries by new lines.
   • Upload File: Supports .txt, .csv, .json, .md files.

3. Analysis:

   • Click "Analyze Feedback" to send data to Gemini.
   • The AI processes the text using a strict JSON schema to ensure structured and reliable output.
   • View the generated "Strategic Action Plan", "Executive Summary", and detailed feedback table.

📂 Project Structure

• App.tsx: Main application controller and layout.
• services/geminiService.ts: Handles interactions with Google's GenAI API. Defines the prompt engineering and JSON schema for the AI.
• components/Dashboard.tsx: Displays analysis results (Charts, Executive Summaries, Strategic Plans).
• components/InputSection.tsx: Handles user input (Text area and File upload).
• components/Documentation.tsx: Internal documentation view.
• types.ts: TypeScript definitions for the AI response schema and application state.

🛡️ Limitations

• Works best with English text.
• Analysis is probabilistic; always review AI recommendations before implementation.
• Performance depends on the length of input text and API latency.