🤖 InsightsFlow: Network & Sentiment Monitor A real-time dashboard that simulates regional network performance and customer sentiment, then uses an agentic AI (NVIDIA Nemotron) to analyze the data and recommend proactive actions. 🚀 Key Features

• Real-Time Simulation: Uses Gemini to generate a realistic, continuous stream of network metrics (latency, loss) and customer feedback (tweets, support calls).
• AI-Powered Perception: Leverages NVIDIA Nemotron to perform real-time analysis on customer feedback, extracting sentiment, topic, and urgency.
• Stateful Happiness Tracking: Calculates short-term and long-term "happiness scores" for each region using a dual moving average to identify trends.
• Agentic Decision-Making: An AI orchestrator analyzes the combined network and happiness data to make proactive decisions, like send_alert or log_and_monitor.
• Live Streamlit Dashboard: A multi-faceted dashboard visualizes all key metrics, including time-series graphs of network health, customer happiness, and a live-updating log of all AI-driven actions. 🛠️ Project Architecture This project runs as four separate, communicating components:
• simulator.py (The World):
  • Generates network/sentiment data.
  • Serves this data at http://localhost:8000.
• agent_listener.py (The Brain):
  • Pulls data from the simulator (:8000).
  • Analyzes it using Nemotron.
  • Makes a decision.
  • POSTs its final report (data + decision) to the Report Server (:8001).
• reporter_with_storage.py (The Cache):
  • Receives and stores the final reports from the AI agent.
  • Serves these stored reports at http://localhost:8001/reports.
• streamlit_dashboard.py (The Dashboard):
  • Fetches data from the Report Server (:8001/reports) every 5 seconds.
  • Visualizes the data for the user. ⚙️ How to Run You will need four separate terminals open to run the full system.

1. Setup First, clone the repository and set up the environment.

Clone this repository

git clone https://github.com/YOUR_USERNAME/YOUR_REPO_NAME.git cd YOUR_REPO_NAME

Create and activate a virtual environment (recommended)

python -m venv venv source venv/bin/activate # On Windows, use: venv\Scripts\activate

Install all required packages

pip install -r requirements.txt

Create your secret .env file

cp .env.example .env

Now, edit the .env file and add your GEMINI_API_KEY and OPENROUTER_API_KEY. 2. Run the System Run each command in a separate terminal, in this specific order. Terminal 1: Start the Data Simulator python simulator.py

(Wait until you see it serving at http://localhost:8000) Terminal 2: Start the Report Server python reporter_with_storage.py

(Wait until you see it listening on http://localhost:8001) Terminal 3: Start the AI Agent python agent_listener.py

(You will see this terminal start polling for data and making decisions) Terminal 4: Start the Streamlit Dashboard streamlit run streamlit_dashboard.py

(This will automatically open the dashboard in your web browser!)