Blog

Inside the SOC

SOC Tiers: What Tier 1, 2 and 3 Analysts Do in an Agentic SOC

SOC tiers divide analyst work into three levels. Tier 1 triages alerts, tier 2 investigates and responds to the alerts tier 1 escalates, and tier 3 hunts threats, builds detections and handles the most complex cases. Under the agentic SOC model, AI agents investigate each alert before an analyst picks it up, so tier 1 analysts review the evidence instead of gathering it, and every tier spends more of the shift making decisions.

What are SOC tiers?

A tiered SOC divides analyst work by how deep the investigation goes. Tier 1 resolves the alerts it can and escalates the rest, and each tier above takes on more complex and specialized work.

Tier

What it's responsible for

Where its work goes

Tier 1

The alert queue. Triages each alert for priority and severity, resolves what it can, documents the findings and escalates alerts that look like real threats.

Likely real threats, escalated to tier 2 with the findings attached

Tier 2

In-depth investigation of escalated alerts. Correlates across data sources, establishes the full scope and leads the response as an incident develops.

Response decisions as an incident develops, plus new detections and threat hunt findings

Tier 3

The most advanced analysis, such as malware analysis, plus complex threat hunting, specialized detection engineering and mentoring newer analysts.

New detections, threat hunt findings and process improvements that change which alerts reach tiers 1 and 2

The boundary between tiers varies from one SOC to another. SANS, for example, lists threat hunting and detection engineering among tier 2 duties, with the most complex and specialized versions at tier 3.

What does a tier 1 SOC analyst do?

A tier 1 SOC analyst is responsible for the alert queue. Each alert gets an initial assessment. The analyst pulls the context that shows whether the alert matters, such as who the user is, what the host normally does and whether the indicator is known to be malicious. Then they decide whether the alert is a real threat or a false positive, and either close it with documented findings or escalate it.

What does a tier 2 SOC analyst do?

A tier 2 analyst takes the alerts tier 1 escalates and determines how far the threat has spread. They correlate across data sources, identify which users and hosts are involved, and lead the response as an incident develops, deciding the next course of action.

What does a tier 3 SOC analyst do?

Tier 3 analysts are the SOC's most experienced specialists. Threat hunters form hypotheses about activity the existing detections might miss, and test them against the data. Detection engineers turn what threat intel and threat hunts surface into new detections. Tier 3 analysts also handle the most advanced analysis, such as malware analysis, and mentor newer analysts.

Do SOCs still need tiers?

Not in any fixed form. SANS notes that the analyst role depends partly on how each organization structures its security team, and that whether a SOC has a tier 3 depends on its size. What matters more than the org chart is that each part of the work has a clear owner on the team, from the alert queue and investigations to threat hunts and detections.

The SANS 2026 SOC Survey, which drew 444 respondents working in security operations, found the same variety in how SOCs handle incident response.

  • Some keep a response team inside the SOC.

  • Some move SOC and non-SOC staff into response roles when an incident calls for it.

  • Others rely on a separate response function.

SANS concludes that no single model is optimal, and that resource availability and incident frequency decide the right one.

Whatever the structure, the work still needs skilled analysts. Lack of skilled staff was the top operational challenge in the same survey, cited by 14% of respondents.

How do SOC tiers change when AI agents investigate alerts?

AI agents change where each tier spends its time. The agents investigate every alert end to end before an analyst sees it, so work reaches each tier with more of the investigation already done. Analysts still own every verdict.

Tier 1: from gathering evidence to reviewing it

For each alert, a tier 1 analyst gathers evidence before deciding.

  • Querying the security information and event management (SIEM) platform

  • Checking the identity provider

  • Looking up the indicator

  • Reading the endpoint timeline

When an AI agent investigates the alert first, it reaches the analyst as a verdict with evidence and a guided recommendation. The analyst reads the agent's reasoning, checks the evidence against what they know about the environment, and owns the call.

Across Dropzone AI deployments, manual alert investigation drops 95% on average. For the steps inside a single investigation, see the alert triage walkthrough.

Tier 2: escalations arrive already investigated

Tier 2 receives the smaller set of alerts that need an analyst's judgment. Each one comes with its evidence chain and a validated plan for response, so the analyst starts with the evidence already assembled.

Analysts set limits on which response actions the agent can take on its own, and widen them as the agent earns their trust. Any action outside those limits goes to an analyst. At Pipe, escalated investigations run 90% faster with Dropzone AI.

Tier 3: threat hunting becomes a regular practice

Thorough threat hunting takes time, expertise and cross-tool investigation, and that's hard to sustain alongside the alert queue. In the SANS 2026 SOC Survey, about 16% of respondents said they don't prioritize threat hunts at all. SANS also notes that when threat hunting is ad hoc and individually driven, its findings rarely feed back into detection engineering or playbook updates in a systematic way.

With Dropzone AI's AI Threat Hunter, the team runs hypothesis-driven threat hunts, and the agent conducts them using 270+ prebuilt hunt packs. At Indiana Farm Bureau, up to 40 hours of threat hunting was compressed to about 1 hour.

Detection engineering and forensics stay with the SOC's analysts and engineers. Threat hunters decide which hypotheses are worth testing, and detection engineers turn what the threat hunts find into new detections.

Tier

Without AI agents

In an agentic SOC

Tier 1

Gathers the evidence for each alert, then decides whether it's a real threat

Reviews the agent's verdict, evidence and guided recommendation, then owns the call

Tier 2

Scopes each escalation by correlating across data sources

Starts from an investigated escalation with a validated plan for response

Tier 3

Fits threat hunts around the alert queue

Directs threat hunts the AI Threat Hunter conducts, and builds detections from what they find

How do you know whether the tier structure is working?

"Number of incidents handled" has been the top reported SOC metric for ten consecutive years, according to the SANS 2026 SOC Survey. This year, 70% of respondents cited it. It measures volume, and says nothing about whether each tier resolves its work faster. SANS also recommends outcome measures, such as time from detection to containment and avoidability rate.

For the tiers, mean time to investigate (MTTI), the span from an analyst or agent picking an alert up to a defensible verdict with a response plan ready, gives a clean measure of analyst throughput, and it's the part of the workflow best placed to benefit from AI agents. For the other SOC time metrics, see the guide to mean time to conclusion.

Key takeaways

  • Tier 1 triages alerts, tier 2 investigates and responds to escalations, and tier 3 hunts threats, builds detections and handles the most complex cases.

  • Whatever the tier structure, what matters is that each part of the work has a clear owner.

  • When AI agents investigate every alert first, tier 1 analysts review verdicts, tier 2 starts from investigated escalations, and analysts set limits on what the agents do on their own.

  • Measure the change with mean time to investigate, not the count of incidents handled.

See how an agentic SOC team is organized

The AI SOC Team Playbook sets out six specialized AI agent roles and how they work as a team, a KPI framework built for AI-augmented operations, and a step-by-step implementation roadmap. It draws on learnings from 300+ deployments.

Download the AI SOC Team Playbook

Frequently asked questions

[

01

]

What are SOC levels?

[

02

]

Is tier 1 SOC analyst an entry-level job?

[

03

]

Will AI replace tier 1 SOC analysts?

[

04

]

Does every SOC need a tier 3?

Sources: SANS 2026 SOC Survey (June 2026; 444 respondents working in security operations), linked in full above.

Table of contents

Subscribe to Our Blog

Share this article
Overlapping translucent violet geometric panels glowing against a dark navy background.

Self-Guided Demo

Test drive our hands-on interactive environment. Experience our AI SOC analyst autonomously investigate security alerts in real-time, just as it would in your SOC.

Overlapping translucent violet geometric panels glowing against a dark navy background.

Self-Guided Demo

Test drive our hands-on interactive environment. Experience our AI SOC analyst autonomously investigate security alerts in real-time, just as it would in your SOC.

Overlapping translucent violet geometric panels glowing against a dark navy background.

Self-Guided Demo

Test drive our hands-on interactive environment. Experience our AI SOC analyst autonomously investigate security alerts in real-time, just as it would in your SOC.

Overlapping translucent violet geometric panels glowing against a dark navy background.

Self-Guided Demo

Test drive our hands-on interactive environment. Experience our AI SOC analyst autonomously investigate security alerts in real-time, just as it would in your SOC.

Subscribe to our newsletter

Get our monthly AI SOC insights and threat insights.

Subscribe to our newsletter

Get our monthly AI SOC insights and threat insights.

Subscribe to our newsletter

Get our monthly AI SOC insights and threat insights.

Subscribe to our newsletter

Get our monthly AI SOC insights and threat insights.