In short
A SIEM (Security Information and Event Management) is a platform that centralizes event logs from the entire information system, standardizes them, and correlates them to identify signs of an attack in the form of alerts. It is the central tool of a SOC: it is used to detect an incident and then reconstruct its timeline using the stored history. Its value depends entirely on the sources collected and the quality of the detection rules.
How Does a SIEM Work?
- Data Collection: logs from servers, the directory service, firewalls, the VPN, email systems, applications, and the cloud.
- Standardization: heterogeneous formats are converted to a common model, with synchronized timestamps.
- Correlation: rules that link multiple events, such as a series of failed login attempts followed by a successful login from an unusual country.
- Alerts and investigation: analysts triage alerts and review the history to understand an attack.
- Retention: a sufficient history to investigate an intrusion discovered weeks later.
Priority sources to connect
- Authentication: Active Directory or identity provider, VPN, email.
- Devices exposed to the Internet: firewalls, gateways, reverse proxies, WAFs.
- EDR alerts on workstations and servers.
- Cloud service administration logs.
- Business application logs: logins, changes to permissions, sensitive actions.
Without application logs, an attack that hijacks an application’s logic remains invisible to the SIEM.
Frequently asked questions
What is the difference between SIEM and SOC?
SIEM is a tool; the SOC is the team and organization that operates it. A SIEM with no one to analyze its alerts offers no protection whatsoever.
Sources
Find out where an attacker would get in.
Our AI agents actually attack your scope, fully automated, and hand you the proof of every vulnerability.
