Home

Failing controls working again, with evidence

ZeroTB investigates failing controls, completes repairs with system owners and checks the results.

Know which repairs are verified
A Cindral control repair: audit logging found off, assigned to the platform lead, the drafted pull request approved, checked again as passing, with its signed evidence record.

ZeroTB carries each failing control to a fix

ZeroTB combines software and a team to repair SOC 2, ISO 27001 and HIPAA controls with system owners at US companies with 25 to 250 people past their first audit.

  • Repair with owners. Our team completes repairs with the people who own the affected systems

  • Cover five surfaces. Five surfaces cover development, cloud infrastructure, identity and access, endpoints, and people and process

  • Connect your systems. Integrations connect GitHub, AWS, Google Workspace, identity providers and endpoints for control checks

Repairs stuck between teams cost audit readiness

A failed control leaves compliance waiting on IT or engineering to repair the system. Another change can bring the same gap back.

The requirement needs someone to carry the repair through to a checked result before the next audit.

  1. A system changes

  2. The control stops working

  3. Teams chase the repair

We take each control from clause to record

Read

Define the operating requirement

We extract the clause into a commitment with scope, owner and checking schedule.

A confirmed commitment linked to its source clause

Keep the original requirement in view
A Cindral commitment drawn from its source clause on audit logging, with its scope, owner, daily check and the security lead's confirmation of its meaning.
See the evidence behind a closed repair
A Verified Control Record for a closed repair: the logging commitment, the state before and after, the action, the approval, the verification and a valid signature.

Monitoring alerts start the repairs ZeroTB finishes

ZeroTB carries a monitoring result through to verified repair. Built-in monitoring in compliance automation platforms identifies where attention is needed.

Your team or provider carries repairs. ZeroTB assigns owners and response deadlines, prepares fixes and checks results before closing.

Your team authorizes the changes ZeroTB prepares

Material changes normally arrive as pull requests for your team's review and merge. Your team also confirms what each commitment means.

  • Start read-only. ZeroTB observes first, before any write access is granted

  • Scope write access. Revocable write access covers specific action classes. API actions are opt-in, reversible and bounded

  • Keep AI advisory. Approved logic runs actions. AI cannot declare a control healthy

Talk to our team

We can discuss where control repairs stall, how approvals work today and what your team needs from the completed repair.

FAQ

Can ZeroTB restore drift through pull requests and runbooks?

ZeroTB restores drift through pull requests, runbooks, approved API actions or guardrails. Pull requests are the default for material changes.

Will ZeroTB fix the failing control, or only report it?

ZeroTB combines software with hands-on service to complete repairs with the system owner. Our team verifies the restored control before closing the work.

What happens if a check loses access to its data?

ZeroTB shows the check as unknown, stale or disconnected. The control stays unconfirmed until there is enough information to assess it.

Who decides whether a control can have an exception?

The people responsible for the commitment decide legitimate exceptions. AI can help interpret the requirement but cannot authorize an exception or declare the control healthy.

Can the same repair support several compliance frameworks?

One repair can support SOC 2, ISO 27001 and HIPAA requirements when they share an underlying control. The record keeps the framework mappings.

How does ZeroTB measure how long restoration takes?

Mean Time to Verified Restoration measures time from detected drift to the verified restored control. Each drift event also has an owner and response deadline.

Start with the Pilot

The eight-week Pilot maps up to 25 commitments and restores two real control gaps, leaving working controls and portable evidence records.