How to Build Your First Torq Security Workflow

November 2, 2026

Torq is a security hyperautomation platform: it connects your EDR, email gateway, IAM and ticketing systems so a single alert can trigger enrichment and response across all of them. A first workflow — phishing report to verdict — takes about two hours to build and immediately removes a repetitive analyst task.

The reason to start there rather than with something impressive: phishing triage is high volume, well-understood, and has a clear right answer most of the time.

What can Torq automate that a SOAR can’t?

Both orchestrate security tools. Torq’s difference is in how workflows get built — it generates steps from natural language and ships a large library of prebuilt integrations, so a workflow is assembled rather than coded. Whether that’s faster depends on complexity. Simple three-step playbooks are genuinely quick; branching logic with 20 conditions is still work.

Pricing is quote-based and scales with workflow volume and seats, which means the business case has to be made in analyst hours saved, not per-ticket.

Step 1: connect three tools, not ten

Integrations → Add for the systems that matter to phishing triage:

  • Email gateway (Microsoft 365 Defender, Google Workspace, Proofpoint) — read the reported message, its headers, any attachments.
  • EDR (CrowdStrike, SentinelOne, Defender for Endpoint) — check whether any user clicked or executed.
  • Your ticketing system (Jira, ServiceNow, Freshservice) — create the case.

Each integration needs a service account with the narrowest scope that works. Read-only for the email gateway unless your workflow also deletes messages.

Step 2: write the workflow as three stages

A phishing triage workflow that holds up:

  1. Trigger — new phishing report from the user’s report button or a shared mailbox.
  2. Enrich — pull sender reputation, URL reputation, attachment hashes, and a list of recipients who received the same message.
  3. Decide — is the verdict malicious (sender domain < 30 days old OR URL on two blocklists OR hash known-bad)? If malicious: purge from all mailboxes, reset the credentials of anyone who entered them, notify the reporters. If not: reply to the reporter with a one-line explanation and close.

Step 3 is where the human boundary should sit. Purging mail company-wide should not be fully automatic in your first month.

What should stay manual at the start?

Four decisions, for the first 30 days:

  • Company-wide mailbox purge.
  • Account disable or credential reset.
  • Blocking a sender domain at the gateway.
  • Anything involving an executive’s mailbox.

Run them as approval steps instead of auto-actions. Torq supports approval gates natively, so this is a toggle per step, not a redesign. Once the workflow’s false-positive rate is known — measured over 50+ alerts — remove approvals from the ones that proved reliable.

How do you measure whether it works?

Two numbers:

Metric Before automation Target
Mean time to verdict 42 min (analyst-sample) under 8 min
Analyst touches per report 3–5 1 (the verdict email)

The second one matters more. If analysts still open every alert to double-check, you’ve added a system without removing work.

Step 3: close the loop with the reporter

Replying to the person who reported the phishing email is the step teams skip. It’s also the one that changes reporting rates. A one-paragraph auto-reply — “we checked this, it was malicious, we’ve removed it from inboxes” — roughly doubles the number of reports within a quarter, because reporting stops feeling like shouting into a void.

Automate the reply text per verdict. Keep it short. Nobody wants a security essay at 9am.

Where Torq doesn’t pay off

Teams receiving fewer than about 50 security alerts a day. Below that volume the setup and integration maintenance cost more analyst time than the automation saves, and a documented runbook does the same job.

Also teams without a reasonably clean tool inventory. If nobody knows which EDR is deployed where, building on top of that inventory is building on sand.

Quick Summary

  • Start with phishing triage — high volume, clear verdicts, contained blast radius.
  • Connect three tools (email gateway, EDR, ticketing) with narrow service accounts.
  • Keep purge, credential reset and domain blocking behind approval gates for the first 30 days.
  • Target under 8 minutes to verdict and one analyst touch per report.
  • Auto-reply to reporters — reporting rates roughly double.
Tags: