Updated by
Hi, I'm Jack, the owner of Cybersecurity Jobs List and co-founder of Himalayas (himalayas.app) and Cavuno (cavuno.com). Across all my platforms, I work with application security daily: dependency vulnerability scanning, secure authentication, API security, and data protection across hundreds of thousands of users. My technical background is in computer science (UNSW). I studied security engineering and computer networks, and worked as a research assistant on VR experiments that were published in the Journal of Experimental Psychology. I also work with cybersecurity hiring data every day, tracking which companies are posting, what certifications actually appear in listings, how salaries differ by sub-discipline and clearance level, and where the talent gaps are widest. That combination of security practice, engineering at scale, and daily immersion in hiring data shapes the content on this site. I'm currently based in Sydney, Australia.

A SOC analyst spends most of the day judging alerts that already exist. A detection engineer spends most of the day deciding which alerts should exist, how they should fire, and how they will be maintained. Both jobs protect the same organization. They are not interchangeable titles.
If you are deciding which path to pursue, or explaining the difference in an interview, start with ownership: who consumes the signal, and who produces it.
What each role owns
| Question | SOC analyst | Detection engineer |
|---|---|---|
| Primary output | Investigations, escalations, case notes, containment steps within policy | Detection content, tests, tuning changes, coverage notes, triage guidance for analysts |
| Typical question | Is this alert real, urgent, and actionable right now? | Should this behavior become an alert, and will analysts trust it at 3 a.m.? |
| Success signal | Correct triage, clear documentation, timely escalation | Useful true positives, controlled false positives, maintainable rules, tested coverage |
| Common tools | SIEM consoles, EDR, case systems, playbooks | SIEM query languages, rule repos, Sigma or native rule formats, test harnesses, CI checks |
| Schedule shape | Often shift-based monitoring and on-call | Often project and backlog work with release discipline; on-call varies by team |
These are patterns, not universal org charts. In a small company one person may wear both hats. In a large SOC the titles may be split across several teams. Read the posting's responsibilities before you assume the label.
How the work differs day to day
SOC analysis is event-driven. You receive an alert, gather context, decide whether it is benign, suspicious, or malicious, and hand off when the case exceeds your authority. Good analysts protect the organization by being precise under time pressure and by refusing to invent certainty the logs do not support.
Detection engineering is product-driven. You take a detection idea (a noisy rule, a missed intrusion technique, a new log source, a purple-team finding), turn it into logic, document expected true positives and false positives, validate it, and ship it so analysts can use it. Good detection engineers protect the organization by making alerts fewer and clearer, not merely more numerous.
NIST's NICE Framework (SP 800-181 Rev. 1) groups related defensive tasks under cybersecurity work roles. Tasks such as developing cyber defense content, constructing signatures, and validating intrusion detection alerts sit close to detection engineering even when employers use different job titles. Use the framework as a vocabulary check, not as proof that every employer organizes work the same way.
A fictional example: from alert to detection
Use this fictional case only as practice. It is not a real incident and not a production rule you should paste into a live SIEM without your team's process.
Observed problem. Analysts keep receiving medium-severity alerts for successful cloud console sign-ins from unfamiliar cities. Most cases close as travel. A few are risky enough that the team wants a sharper signal when a new or rarely used account signs in without multi-factor authentication.
SOC analyst path. On each alert, the analyst checks account owner, device, MFA status, travel tickets, VPN history, and recent permission changes. They document observed facts, a hypothesis, unknowns, and the next owner. They may map candidate behavior to MITRE ATT&CK Valid Accounts (T1078) when the evidence supports account abuse, while keeping that mapping separate from proof of compromise. Our ATT&CK mapping guide walks through that discipline.
Detection engineer path. After several shifts show the same weak alert, a detection engineer frames a narrower detection goal: successful console login without MFA for accounts created in the last N days, excluding known break-glass automation. They choose the log source fields, write the rule, list expected false positives (new hires on day one, approved admin workflows), define a test case, and write short triage guidance so the next analyst knows which fields matter first.
A portable starting point is a Sigma-style rule skeleton. Sigma is a common open format for describing detections before they are translated into a specific SIEM language. The example below is simplified training material, not a complete production detection:
title: New Cloud Account Console Login Without MFA
status: experimental
description: Fictional training example. Successful console login without MFA for recently created accounts.
logsource:
product: cloud
service: signin
detection:
selection:
EventName: ConsoleLogin
ConsoleLogin: Success
MfaUsed: "No"
filter_known_break_glass:
UserIdentityType: AssumedRole
UserIdentityArn|contains: break-glass
condition: selection and not filter_known_break_glass
falsepositives:
- New employee first login during onboarding
- Approved emergency access workflows not yet tagged
level: medium
Before anything like this reaches production, a real team would map exact field names for their cloud provider, add environment-specific exclusions, test with known good and known bad events, and document how analysts should escalate. David Bianco's Pyramid of Pain is a useful reminder that detections aimed only at disposable indicators (hashes, IPs) are easier for an adversary to bypass than detections aimed at behaviors and procedures. The pyramid is a teaching model, not a scoring system for your resume.
Skills that transfer and skills you must add
From SOC into detection engineering, you already bring:
- Pattern recognition for real attacks versus routine noise
- Comfort reading logs under incomplete evidence
- A practical sense of which alert fields help an investigation
- Experience writing clear case notes other people can continue
You usually still need to build:
- Query and rule fluency in your organization's SIEM language
- Version control habits (reviewable changes, not one-off console edits)
- Testing discipline: prove the rule fires on the intended behavior and stays quiet on known benign cases
- Product thinking: backlog, ownership, rollback, and triage notes for the SOC as the customer
From software engineering into detection engineering, you already bring automation and review habits. You usually still need attacker and SOC context so you do not ship high-volume alerts nobody can investigate.
Neither path requires you to invent a personal malware lab story. Employers care whether you can explain a detection goal, a false-positive risk, and a validation method.
How to read job postings
Look past the title. Some "SOC analyst L3" roles are mostly detection content. Some "detection engineer" roles are mostly advanced triage with light rule edits. Use this checklist:
- Does the posting emphasize shift work, ticket queues, and incident playbooks, or rule repositories, coverage metrics, and purple-team validation?
- Are you measured on queue time and case quality, or on detection precision and maintainability?
- Is coding required for production changes, or only helpful for ad hoc searches?
- Who is named as the customer of your work: leadership reporting, end users, or the SOC analyst team?
When you apply, bring one concrete example. A strong SOC example is a case where weak evidence forced a careful escalation. A strong detection example is a noisy alert you helped redesign, with before/after false-positive notes and no confidential data.
Portfolio practice without live production access
If you do not yet own production rules, you can still practice safely:
- Pick one public technique page, such as T1078, and write the behavior in plain language first.
- List the log sources that would be required to see that behavior in a typical cloud or endpoint environment.
- Draft a Sigma-style detection with explicit false positives and unknowns.
- Write five triage questions an analyst should answer before escalating.
- State how you would test the idea without attacking a system you do not own.
Keep the packet labeled as a training exercise. Do not claim it was deployed. Pair it with interview practice from the SOC interview guide and role-fit notes from SOC analyst versus incident responder.
Which role should you choose?
Choose SOC analysis if you want real-time investigation skill, broader exposure to many alert types, and a common entry path into security operations. Choose detection engineering if you prefer building and maintaining the detection product, can protect focus time for engineering work, and enjoy making other analysts faster.
Many people do SOC work first, then move into detection once they can show repeated examples of noisy or missing alerts they helped fix. That sequence is common because the hardest part of detection quality is knowing what a useless alert feels like to the person who has to live with it.
Browse live openings on the job board, then read each description for ownership, schedule, and validation expectations before you decide the title matches the work you want.


