Aurora AI Security

Overview

Aurora AI Security is my Masterschool capstone project — a four-phase lab that simulates the security infrastructure of a fictional AI company called Aurora AI Security GmbH. The goal is to design, deploy, and validate an enterprise-grade security environment from scratch.

Project Phases

Phase 1 — Network Planning (Completed)

Designed a VLAN-segmented network (192.168.64.0/24) mapped to the NIST Cybersecurity Framework 2.0. Deliverables included:

• Network topology diagram with DMZ, internal subnets, and management VLAN
• IP addressing scheme and DHCP scoping
• Risk assessment aligned to GDPR and NIS2 requirements
• Incident response workflow design

Phase 2 — Virtual Infrastructure (In Progress)

Standing up eight virtual machines on Apple M3 via UTM (QEMU):

• pfSense CE — perimeter firewall with VLAN routing
• Alpine Linux — virtual switch fabric
• Wazuh 4.x (Manager + Indexer + Dashboard) — SIEM stack
• Ubuntu Server — Wazuh host
• Windows 11 ARM — endpoint agent
• Kali Linux — attacker simulation

Phase 3 — Security & SOC Planning (Upcoming)

Threat modeling using MITRE ATT&CK, detection rule development, and playbook authoring for common incident scenarios.

Phase 4 — Implementation & Validation (Final)

Penetration testing of the lab environment, vulnerability scanning with OpenVAS/Nessus, and a full after-action report.

Key Technical Decisions

• Chose Wazuh over a hosted SIEM for cost and hands-on experience with agent deployment and rule tuning
• Used pfSense because it closely mirrors enterprise firewall products and has strong documentation
• Attack simulation via Atomic Red Team rather than manual exploitation for reproducibility

Skills Demonstrated

• SIEM deployment and configuration
• Firewall rule management
• Network segmentation and VLAN design
• Compliance framework mapping (NIST, GDPR, NIS2)
• Threat detection and log analysis