Skip to main content

Finding Your Way Around Torq

Understand what each Torq product area does.

Overview

Torq is an AI SOC platform built around core product areas that operate as one continuous system. Alert intake, investigation, and automated response flow end-to-end across your tool stack. This article covers each product area and what it does.

Auto Triage

Auto Triage is Torq’s agentic triage engine. It starts working once your alert sources are connected. No workflows to build, no custom code required. It applies threat intelligence, business context, and historical case data to every incoming alert. This includes the role, access, and risk posture of any involved entity.

Each alert gets a full investigation report: a verdict, severity, justification, MITRE ATT&CK mapping, supporting evidence, and suggested response steps. This is the analysis a Tier 1 analyst would perform manually, applied to every alert.

What happens after a verdict depends on how the alert source is configured. A true positive can automatically open a case, trigger a workflow, or stay in the queue for manual review. The analyst controls that configuration. Nothing advances beyond what they have defined.

When an analyst corrects a verdict, the correction can be saved as a permanent rule. Future alerts matching the same pattern automatically follow that rule. This is how Auto Triage gets more accurate over time.

Case Management

A case is the investigative workspace where a threat is tracked from first evidence through resolution. Cases can be created from Auto Triage true positives, through workflows, or manually.

Each case surfaces status, severity, SLA, and threat category. The AI-generated summary covers what happened, when it happened, the potential impact, and the key indicators involved. Any analyst stepping in understands the case immediately. The Timeline records every case action in chronological order. Analysts can trigger workflows from within a case.

When a case is created via Auto Triage, every observable is enriched automatically. Each indicator receives threat-intelligence context and a verdict, and is tracked across all cases. The next time an indicator appears, every analyst sees the prior decision and who made it.

AI Agents

AI Agents are custom intelligence units you define for specific SOC tasks. Each agent has a role, instructions, tools, and parameters. Build from templates, from scratch, or through Socrates Builder. Torq ships pre-built AI Agents ready to deploy.

Once published, AI Agents are available across the platform: as steps in workflows, through Socrates, and in Case Management. Every agent action is governed by Guardrails.

Socrates

Socrates is Torq’s AI orchestrator. It runs investigations, builds workflows, and directs AI Agents across the platform. Cases can be assigned to Socrates the same way they are assigned to any analyst. Analysts interact with Socrates through three modes.

  • Socrates Builder: The recommended path for creating AI Agents and building workflows. Describe what you need, and Socrates Builder plans, generates, and validates it. Nothing goes live until you approve it. It also handles workspace-wide changes. To learn more, see: Socrates Builder: Build Workflows and AI Agents

  • Socrates Investigator: Investigate any signal across your workspace without opening a case. Socrates retains access to all your data, tools, and AI Agents outside a case. Use it to research indicators, hunt for similar activity, generate detection rules, or compile reports. Sessions are saved on the Socrates Conversations page. To learn more, see: Socrates Investigator: Investigate Threats Outside a Case

  • Chat with AI Agents: Select a published AI Agent from the Socrates picker and start a conversation. To learn more, see: Socrates: Chat with AI Agents

Every step Socrates takes is visible in real time. Every AI Agent it calls is fully audited. Inputs, outputs, and decisions are open for review.

Hyperautomation

Hyperautomation is Torq’s workflow automation layer. Every workflow uses a trigger, steps, and operators to execute security processes across your tool stack.

Triggers start a workflow from an alert, a case event, a schedule, or on demand. Steps are the building blocks: API calls, data transformations, case actions, and integrations with third-party tools. Find and drag steps directly from the Builderbox. Operators control logic and flow. Utilities handle data extraction, transformation, and organization within the workflow. Torq ships 300+ pre-built integrations and 4,000+ pre-built steps. Templates give you ready-made starting points for common SOC use cases.

You can also use Socrates Builder to generate a complete workflow from a plain-language description, then refine it manually.

Did this answer your question?