Video walkthrough available
Project 01 · Defender XDR · Entra ID · Email → Identity → Endpoint

Multi-Stage Phishing → Account Compromise → Malware Execution & Endpoint Response

A controlled lab investigation showing how I used Defender XDR, Defender for Office 365, Defender for Endpoint, Entra ID, and KQL to connect phishing evidence, suspicious identity activity, malware execution, and endpoint response decisions.

Defender XDRDefender for Office 365Defender for EndpointEntra IDMalwareKQLAWS EC2
CONFIRMSCOPECONTAININVESTIGATEERADICATERECOVERDOCUMENTESCALATE
Scenario

Controlled multi-stage investigation

The scenario begins with a controlled phishing simulation delivered to lab users. It is designed to connect user interaction with identity-related suspicious activity and endpoint follow-on analysis. The recording and 19 supporting screenshots document the incident workflow, hunting queries, and response interfaces used in the lab.

Environment

Telemetry available to investigate

Microsoft Defender XDR, Defender for Office 365, Defender for Endpoint, Entra ID, Intune, Azure/Azure Arc, AWS EC2 Windows Server endpoints, KQL Advanced Hunting, PowerShell, and supporting OSINT/sandbox tools. This is a technology inventory, not a claim that every product is directly integrated.

Video Evidence

Watch the investigation walkthrough.

A full narrated walkthrough explaining the incident, investigation pivots, response approach, and conclusions. Supporting screenshots are organized by stage below.

Project 01 · Investigation Walkthrough10:31 · Controlled SOC lab · Recorded evidence
Download video
Jump to a sectionFull recording · Original sequence
01 · CONFIRM / VALIDATE

Determine whether the alert represents real risk.

Review the phishing message, sender and authentication context, message trace, URLs or attachments, reported user interaction, related Defender alerts, and initial identity/endpoint signals. Separate confirmed facts from hypotheses.

01 · Incident accepted for investigation

Manage incident panel showing the In Progress status and incident classification.

02 · Alerts and incident graph

Defender XDR alert chain and relationships between incident entities.

02 · SCOPE

Establish the full blast radius.

Pivot across recipients, clicks, affected accounts, sign-ins, sessions, mailbox rules, devices, processes, files, network connections, and related indicators. Build a timestamped incident timeline and identify what is affected versus merely exposed.

03 · Affected assets

Assets associated with the incident, used to scope the endpoint investigation.

07 · UrlClickEvents: user interaction

URL-click query and results used to review which accounts interacted with links.

11 · Entra ID sign-in review

Sign-in log view and time filter. No sign-in records are displayed in this capture.

03 · CONTAIN

Stop active risk while preserving evidence.

Select containment proportional to confirmed activity: remove malicious mail, block validated indicators, revoke suspicious sessions, restrict the affected identity, or isolate an endpoint when endpoint compromise is supported.

12 · Password reset

Password-reset panel for the lab identity.

13 · Require MFA re-registration

Confirmation prompt for requiring the user to register MFA again.

14 · Revoke sessions

Confirmation notification for revoking the user’s sessions.

15 · Tenant IP blocking

Tenant Allow/Block List IP-entry panel, before a block is submitted.

16 · File-hash indicator

File hash entered in the Add indicator form.

18 · Aria device isolation

Device-isolation confirmation prompt for the Aria endpoint.

04 · INVESTIGATE / ROOT CAUSE

Correlate email, identity, and endpoint evidence.

Use Defender XDR and KQL to connect the delivery and interaction timeline with sign-in context, MFA events, mailbox activity, process trees, command lines, files, and outbound connections. Validate the likely entry point and follow-on behavior without treating impossible travel alone as proof.

04 · PowerShell alert

PowerShell alert details and command context reviewed during endpoint analysis.

05 · PowerShell entities

Entity details associated with the PowerShell alert.

06 · EmailEvents: email delivery

Advanced Hunting query and displayed email-event results.

08 · EmailAttachmentInfo: attachments

Attachment query and displayed results for the email investigation.

09 · DeviceFileEvents: file activity

Endpoint file-event query and displayed results.

10 · DeviceNetworkEvents: connections

Network-event query and displayed results for endpoint correlation.

19 · Aria Live Response

Live Response workspace for Aria, with the session in Waiting state.

05 · ERADICATE

Remove the verified cause and persistence.

Remove confirmed malicious artifacts or unauthorized access paths. Check inbox rules and forwarding, suspicious apps/sessions, startup locations, registry run keys, scheduled tasks, and other persistence relevant to the evidence.

17 · Aria antivirus scan

Antivirus scan prompt for the Aria device; scan completion is not shown in this frame.

06 · RECOVER

Restored access and verified normal activity.

I re-scanned the affected endpoints, reviewed sign-ins and mailbox activity, checked for recurring persistence and suspicious outbound connections, and restored access in a controlled way. I continued monitoring for signs of recurrence.

07 · DOCUMENT

Documented the findings and response decisions.

I documented the investigation timeline, affected assets, evidence, hunting queries, containment and remediation actions, and recovery checks. The record separated confirmed observations from assumptions and captured the reasoning behind my decisions, remaining limitations, and lessons learned.

08 · ESCALATION CRITERIA

Recognize when specialist support is needed.

I would escalate evidence of widespread compromise, privileged-account abuse, persistent access, or suspected data exfiltration to the incident response or DFIR team. My handoff would include confirmed findings, affected assets, the timeline, actions taken, and the specific decision or support required. IT would support restoration and coordinate operational impact.

Lab disclosure: This case study represents controlled home-lab work. Screenshots show the recorded interface state. Action prompts document the response workflow; completed outcomes are identified only where a result is visible.

Investigation screenshot

Open original image ↗