Most SOC 2 checklists read like they were written for a company that already has a security team. You don't. So this is the version for a founder or engineer doing it without a compliance hire. For each item: what "done" looks like, and what a tool can verify automatically versus what needs a human.

One reframe: SOC 2 Security isn't features you buy. It's proof that a handful of sensible practices run consistently and that you can show evidence. Most you may already do informally.

1. MFA everywhere

Done: MFA enforced on every account reaching sensitive systems (identity provider, cloud consoles, code repos, critical SaaS admin). No standing exceptions for founders. Automatable: whether MFA is enforced, read from those systems and monitored. Human: deciding scope and edge cases.

2. Centralized logging

Done: security-relevant events flow to one place, retained for a defined period, with someone who'd notice an alert. Automatable: confirming logging is on and retention configured. Human: defining what you monitor for and who responds.

3. Encryption in transit and at rest

Done: TLS everywhere public-facing; storage and databases encrypted at rest (a setting on AWS/Azure/GCP). Automatable: checking encryption and TLS against live config. Human: key-management approach and documentation.

4. A patch and vulnerability process

Done: a repeatable way to find vulnerabilities and a followed timeline to fix by severity. Automatable: whether scanning is on, endpoints updating, open findings tracked. Human: triage and prioritization judgment.

5. Documented vendor reviews

Done: a list of third parties that touch your data, a quick risk assessment of each, evidence you check their posture. Automatable: the inventory and review dates. Human: the risk judgment.

6. Written policies your staff acknowledge

Done: core policies (access, acceptable use, incident response, data handling) in writing, with a record every employee accepted them. Automatable: tracking acknowledgements. Human: writing policies that reflect how you actually operate. The auditor checks reality against the document.

7. Access reviews

Done: on a cadence (quarterly is common), someone confirms each person still needs their access; access removed promptly on departure. Automatable: pulling the who-has-access picture and flagging departed staff. Human: the review and sign-off.

8. Background checks

Done: a documented practice of appropriate pre-employment checks, applied consistently, within local law. Automatable: tracking a check was completed. Human: everything else. This is a people process.

9. A basic risk assessment

Done: a written, periodically updated assessment of the risks to your data, likelihood and severity, and what you're doing about them. Real and revisited, not 40 pages. Automatable: reminders and linking risks to controls. Human: the thinking. Only you know your actual risks.

The pattern

Software is good at the continuous checking (is MFA on, is encryption enabled, who has access right now) and at keeping time-stamped evidence. Humans own the judgment and documentation. That is what continuous control monitoring does: checks run against your live systems on an ongoing basis, and the evidence is already there.

Handling this without a hire

This is the gap Argyx was built for: getting SOC 2 ready without a security or GRC hire. It connects to your identity provider, cloud, and endpoints to test the automatable items continuously and auto-collect evidence, and it tracks the human items so you can see what's done, what's due, and what's missing. At audit time, your auditor gets a read-only login. Pricing starts at $299/month.

The honest caveat: software can't write your risk assessment or make the judgment calls, and any tool that claims otherwise is overselling.

If you want to see where you stand against this list today, run a free scan of your live systems (no login required), or start the 14-day free trial.