NetPulse AI

Developed by Team HELIX AI

This project monitors network health using a Raspberry Pi, collecting data on CPU usage, temperature, signal strength, and packet loss. The data is logged in Excel, identifying abnormal conditions. Users upload the data to a Hugging Face chatbot for predictive analysis and optimization recommendations.

Features:

  • Future Performance Prediction: Identifies upcoming failure risks based on past data trends.
  • Risk Analysis and Potential Issues: Detects high-risk periods (e.g., peak CPU load at specific hours). Identifies network bottlenecks, low signal strength, or overheating components.
  • Preventive Actions and Recommendations: Suggests cooling measures if temperature spikes are detected. Recommends bandwidth optimization if packet loss is increasing. Alerts users about critical failures and suggests preventive maintenance. Libraries Used:
  • Gradio: For creating the user interface.
  • Pandas: For reading and analyzing Excel files.
  • Hugging API and LLM: Zephyr-7b-beta For utilizing state-of-the-art language models.

How It Works:

  • Detailed Analysis
  1. Upload your data in form of excel file.
  2. Paste it in the Detailed Analysis Tab.
  3. Get detailed recommendations!
  • General Chat for Network Optimization
  1. Talk to AI for more suggestions and queries regarding Network Issues.