Built an agentic SOC investigation platform for Splunk Enterprise using LangGraph, FastAPI, React, and RAG architectures. The system transforms security alerts into explainable investigations, MITRE ATT&CK mappings, containment plans, and detection coverage analysis. Multiple AI agents collaborate to classify threats, reconstruct attack paths, generate reports, and identify detection gaps. Key features: • Multi-agent investigations • MITRE ATT&CK mapping • Splunk integration • RAG-powered analysis • Automated reporting