
This project is an AI-powered multi-agent Network & Security Incident Response platform built using Streamlit, HuggingFace LLMs, LangChain, ChromaDB, and AMD GPU infrastructure. The system ingests network and security logs, classifies them into relevant domains, and routes them to specialized AI agents for network troubleshooting, security analysis, and Root Cause Analysis (RCA). Using Retrieval-Augmented Generation (RAG), semantic log understanding, and multi-agent orchestration, the platform correlates events such as BGP flapping, OSPF instability, packet drops, firewall anomalies, and CPU spikes to generate intelligent troubleshooting insights, remediation steps, and verification commands in real time.
10 May 2026

Agentic AI-Powered SOC Dashboard for Real-Time Threat Detection This project is a cloud-based Security Operations Center (SOC) dashboard designed to perform real-time threat detection and analysis for IP addresses and domain names. The system uses intelligent rule-based logic to classify inputs into low, medium, or high-risk categories based on patterns, keywords, and known safe entities. It is built using a Flask backend deployed on Google Cloud Run, with BigQuery used for storing and querying transaction logs. The dashboard provides live monitoring features including a dynamic alerts panel, API usage analytics, KPI metrics, and a paginated transaction table. The UI is designed in a SOC-style format to simulate real-world security tools like Splunk. The platform demonstrates how AI-driven decision logic, cloud infrastructure, and real-time data visualization can be combined to build a scalable and practical cybersecurity monitoring solution.
26 Apr 2026