Generative AI in SOCs: incident response, human-in-the-loop design, and the balance between acceleration and responsibility.

Dataset, stimulus images, and Grad-CAM visualizations from the research paper: A (Mid)journey Through Reality.

A shows bottlenecks in human only workflows while B is for agentic and HITL workflows to ensure accountability and to prevent automation bias

This is a workflow for an AI hiring portal with governance that achieves regulatory compliance through HITL

Semantic & Governance Layer for agent communication — a semantic + governance overlay on top of A2A/MCP that normalizes meaning, infers resource types, and enforces calibrated trust, approval, and audit. Emergency Bridge MVP (CAP to EDXL-RM).