Threat Hunting, SIEM Detection and Digital Forensics Basics
Write and tune SIEM detection rules, run a hypothesis-driven hunt, triage an endpoint, preserve evidence, build a timeline and write the incident report.
What you will learn
- Map common investigation questions to the right log sources and fix time and coverage gaps.
- Write, test and tune a Sigma detection rule and turn it into a SIEM use case.
- Map detections and incidents to MITRE ATT&CK and read coverage gaps correctly.
- Plan and run a hypothesis-driven threat hunt and turn its result into a lasting detection.
- Triage an alert and run a first endpoint check using Windows artefacts and process evidence.
- Capture memory and disk evidence in the right order and keep a defensible chain of custody.
- Build a UTC timeline from several sources and write a clear, evidence-based incident report.
Modules
- Log sources and visibility
- SIEM use cases and Sigma rules
- Mapping to MITRE ATT&CK
- Hypothesis-driven threat hunting
- Alert triage and endpoint forensics basics
- Memory forensics, evidence handling and chain of custody
- Timelines and incident reports
Course facts
- Estimated study time
- about 60 minutes
- Final assessment
- 42 questions · Pass mark 80%
- Available in
- English, Français, العربية
- Price
- Free. Optional verifiable certificate: $19
- Last updated
- Published by
- AMAADOR ACADEMY
Who this course is for
SOC analysts, junior incident responders, IT and network administrators and security engineers who want to move into detection engineering, threat hunting or DFIR roles.
Prerequisites
Basic networking (IP, DNS, TCP/UDP), everyday Windows and Linux use, and comfort reading log lines. No prior security certification is needed.
Course content
1. Log sources and visibility
Every investigation starts with five questions: who signed in, what ran, where it connected, what changed, and what left the network. Each question has a log source that answers it best. If you do not collect that source, you cannot answer the question later.
| Question | Best source | Useful IDs |
|---|---|---|
| Who signed in? | Windows Security log, identity provider sign-in logs | 4624 success, 4625 failure, 4672 admin rights |
| What ran? | Process creation with command line, Sysmon, PowerShell logs | 4688, Sysmon 1, PowerShell 4104 |
| Where did it connect? | Firewall, proxy, DNS, Sysmon network events, Zeek | Sysmon 3 and 22 |
| What changed? | Security and System logs, Sysmon registry events | 4720 new account, 4732 group change, 7045 new service, Sysmon 13 |
| Was evidence erased? | Security log | 1102 audit log cleared |
Two settings are often missing by default. Process creation events (4688) only include the command line if you enable the policy Include command line in process creation events. PowerShell Script Block Logging (event 4104) must be enabled by policy. Sysmon from Sysinternals is not installed by default: you deploy it with a configuration file you maintain.
2. SIEM use cases and Sigma rules
Sigma is an open, vendor-neutral format for detection rules written in YAML. You write the logic once, then convert it for your SIEM with tools such as sigma-cli and pySigma, which have backends for Splunk, Microsoft Sentinel (KQL), Elastic and others. Only three parts are required: title, logsource and detection. The rest is metadata that makes a rule reviewable.
| Field | Example for encoded PowerShell |
|---|---|
| title | Suspicious encoded PowerShell command |
| id | a random UUID (version 4) |
| status | test (also: experimental, stable, deprecated, unsupported) |
| tags | attack.execution, attack.t1059.001 |
| logsource | category: process_creation, product: windows |
| detection | selection: Image|endswith: \powershell.exe and CommandLine|contains: -enc; filter: ParentImage|endswith: \approved_tool.exe; condition: selection and not filter |
| falsepositives | Admin scripts from the deployment tool |
| level | medium (also: informational, low, high, critical) |
Fields inside one selection combine with AND. A list of values for one field combines with OR. Modifiers after the pipe change the match: contains, startswith, endswith, all (every value must match) and re (regular expression). Tags use lowercase names, such as attack.t1059.001 for the ATT&CK sub-technique.
3. Mapping to MITRE ATT&CK
MITRE ATT&CK is a public knowledge base of adversary behaviour built from real observations. At the time of writing the current release is v20 (13 October 2026). Its Enterprise matrix lists 15 tactics, 227 techniques and 479 sub-techniques, and MITRE publishes major releases twice a year. Always check the version your tools use, because tactic names and IDs change between releases. Since v19 (April 2026) the old Defense Evasion tactic has been split into Stealth and Defense Impairment, so older material that says Defense Evasion may be out of date.
Read ATT&CK from left to right: goal, method, variant, real example.
| ID | Name | Tactic |
|---|---|---|
| T1566.001 | Phishing: Spearphishing Attachment | Initial Access |
| T1059.001 | Command and Scripting Interpreter: PowerShell | Execution |
| T1003.001 | OS Credential Dumping: LSASS Memory | Credential Access |
| T1021.001 | Remote Services: Remote Desktop Protocol | Lateral Movement |
| T1486 | Data Encrypted for Impact | Impact |
ATT&CK also links techniques to detection strategies and data components, which suggest what to log and what to look for. In the v20 Enterprise release MITRE lists 703 detection strategies and 106 data components. Use them as a starting checklist, not as a complete answer.
4. Hypothesis-driven threat hunting
Detection rules wait for a known pattern. Threat hunting is a human-led search for attacker behaviour that no alert has caught yet. It only works when each hunt starts from a clear hypothesis, not from browsing logs. Published frameworks such as PEAK from Splunk and TaHiTI structure this work, and all of them share the same idea: prepare, execute, then act on what you learned.
A good hypothesis names the behaviour, the technique and the evidence you expect:
If an attacker with stolen credentials moves between workstations over RDP (T1021.001), we will see Windows logon events 4624 with logon type 10 from one workstation to another. Normal users in this network do not do this.
- Pick the trigger — threat intelligence about your sector, a new ATT&CK technique, an incident lesson or an anomaly you noticed.
- Write the hypothesis — behaviour, technique ID, expected evidence, and the log source that holds it.
- Check the data — confirm the log exists for the scope you chose and covers the time window. If not, your first finding is a visibility gap.
- Define scope and time-box — for example workstations in one office, last 30 days, two analyst-days.
- Decide what proves or disproves — write it down before you query, so you do not bend the result later.
5. Alert triage and endpoint forensics basics
Triage decides whether an alert deserves an investigation. A routine keeps you fast and consistent when the queue is long. Your goal is a clear verdict backed by evidence, not a long essay.
- Read the rule — what behaviour was it written to catch, and what does the alert actually show?
- Identify the entities — user, host, process, file hash, IP address and time, all in UTC.
- Enrich — who owns the host? Is the user an admin? Is the IP or hash known in your threat intelligence? Was there a change ticket?
- Look around the event — what ran just before and after, on this host and from this user?
- Decide — true positive, benign true positive (real behaviour, approved), or false positive. Escalate true positives with your evidence.
Do
- Write the verdict and reason in the ticket every time.
- Escalate when impact is unclear and the asset is critical.
- Feed false positives back to the rule owner.
Don't
- Close an alert because the severity looks low.
- Close an alert because the host owner says it is fine, without evidence.
- Reboot or wipe the host before evidence is collected.
6. Memory forensics, evidence handling and chain of custody
RFC 3227 tells collectors to proceed from the most volatile evidence to the least volatile. In practice: CPU registers and caches, then routing and ARP tables, the process table and memory, then temporary files, disk, remote logs and finally archive media. Memory holds things that vanish at shutdown: running and hidden processes, network connections, decrypted data, injected code and sometimes credentials.
- Acquire memory first — use a trusted tool such as WinPmem, Magnet RAM Capture or FTK Imager, run from external media, and write the dump to external storage, not to the suspect disk.
- Hash and record — calculate a SHA-256 hash of the dump, then note tool, version, time in UTC and who ran it.
- Analyse a copy — use Volatility 3 (the 2.28 series at the time of writing) on a working copy. It downloads symbol tables for Windows automatically when it has internet access.
| Plugin | Use it to |
|---|---|
| windows.pslist and windows.pstree | List processes and their parent-child relationships |
| windows.psscan | Find hidden or terminated processes missing from pslist |
| windows.cmdline | See each process's command line |
| windows.netscan | Find network connections and listeners |
| windows.malfind | Flag memory regions that look like injected code |
7. Timelines and incident reports
A timeline puts events from different systems into one ordered story. It shows the order of attacker actions, the time between them and where you have no evidence. Without a timeline you collect facts; with one you can explain what happened.
- Collect — SIEM events, endpoint artefacts, firewall and proxy logs, identity logs, email and cloud audit logs.
- Normalise time — convert everything to UTC and write down each source's offset and any known clock drift.
- Merge and sort — use a spreadsheet for small cases or tools such as Plaso (log2timeline) with Timesketch for large ones.
- Annotate — add the evidence reference, the ATT&CK technique and a confidence level to each key row.
- Find the edges — identify the earliest known malicious event and the last one. Keep searching either side.
| Time (UTC) | Source | Event | Technique |
|---|---|---|---|
| 08:12:44 | Mail gateway | Invoice.docm delivered to user A | T1566.001 |
| 08:31:09 | Sysmon 1 | winword.exe starts powershell.exe -enc | T1059.001 |
| 08:31:41 | Proxy | First request to a newly registered domain | T1071 |
| 09:05:20 | Security 7045 | New service installed | T1543.003 |
The full lesson continues in the interactive course on this page.
Sources and references
- MITRE ATT&CK knowledge base (Enterprise matrix, v20)
- Sigma rules: structure and basics
- SP 800-61 Rev. 3: Incident Response Recommendations and Considerations for Cybersecurity Risk Management
- SP 800-86: Guide to Integrating Forensic Techniques into Incident Response
- RFC 3227: Guidelines for Evidence Collection and Archiving
- ISO/IEC 27037: Guidelines for identification, collection, acquisition and preservation of digital evidence
- Sysmon (Sysinternals) documentation
- Volatility 3 documentation
Professional training course for education only. It is not legal advice and does not replace your organisation's incident response plan, your legal counsel or a certified forensic examiner. Rules on collecting and presenting digital evidence differ by country: follow local law and your employer's policy, and only investigate systems you are authorised to examine. Tool names and versions were checked in November 2026; re-check the vendors' documentation before you rely on a specific command.
All courses · Cybersecurity and IT — Free courses, certificate from $7