Questions about VeriProof? Start here.

It puts AI agents into production and keeps every one of them under control. Each consequential action is checked against the rules for its application while your application can still change it. Exceptions go to whoever holds the authority to decide. Every decision is mapped to the frameworks you answer to and leaves a record your compliance team, auditors, and board can read without an engineer. Your application still carries out the action, and each agent you add joins the same portals, handoffs, and records. See an agent at work →

No. Plenty of good tools check a rule before an agent acts, and they serve developers well. VeriProof also gives each person around the agent a portal of their own: Business Owner, Compliance Officer, Governance Engineer, Developer, Auditor, and Administrator. Work moves between them through built-in handoffs, so a governance engineer's change to a production agent waits for an administrator or compliance officer, and a risk decision lands with the business owner. Add records in your industry's vocabulary and evidence an outside reviewer can verify, and you have what gets agents out of pilot and keeps them manageable as you add more. See the roles and handoffs →

It works alongside them. VeriProof supplies policy results, review history, and decision evidence from your AI workflows for use in your existing governance process. Explore integrations →

Keep those tools. VeriProof adds the business-policy and review context behind AI actions, which can support investigations and operational monitoring.

No. VeriProof governs the agents you connect to it. It doesn't discover or inventory unknown agents across your organization. If you need that, keep your discovery tool: VeriProof works alongside it, and you can connect the agents it finds when you want them under your rules.

Consider a payment, an eligibility decision, or a customer response. VeriProof fits workflows where the business needs control over AI actions and a record it can review. Browse industries →

A full-feature pre-production evaluation, free for 60 days from successful account setup. Users and agents are unlimited. All enterprise software features are included, including evidence exports, human review, AI assists, and OAuth sign-in. The limits are 1,000 recorded decisions per month (calendar months in UTC) and 14-day retention. Verify your work email to start without staff approval or payment details. Start free for 60 days →

All account functionality stops, including viewing records and exporting evidence. Contact VeriProof for an extension or to choose a plan in the same account. Records keep their normal retention deadlines. Upgrading cannot restore records already removed.

Begin with the person who owns the business workflow, a developer who understands the application, and whoever will have to sign off on production. Bring in other specialists as their questions arise.

Use your work email, name, and company to create an account. You must be 18 or older. Verify the emailed code, then open your account and sign in. No payment details or staff approval are required. Start with test data and one decision your team wants to govern.

Look at a routine action and an exception that needs review. Can your team understand the result and find the evidence it needs? That gives you a practical basis for judging fit.

Yes. Request implementation help for a specific integration, policy, review path, or test. It is an optional paid service, subject to staff availability. Developer remains free for 60 days; the service does not change its limits or duration.

The written scope can include a tested connection, configured policies and review paths, documented test results, and a handover. VeriProof confirms staffing, responsibilities, timing, and fees before work starts. Sending a request does not authorize a charge or start a subscription.

Plans are sized by agents and recorded decisions. Your rules, vocabulary, and review paths carry to each agent you add. The options are Developer, Production Control, Portfolio Control, and Private Enterprise; production subscriptions use annual contracts and invoice billing. Review pricing →

Compare the plans, then contact VeriProof to agree the plan and subscription terms. Staff apply the change in your existing account. Your configuration and records still within retention remain; upgrading cannot restore records already removed. Each paid plan has its own features and allowances. Choose Portfolio Control if you need portfolio questions, legal hold, custom roles, or external-auditor links. Implementation help is a separate service and does not extend Developer access.

Supported SDKs, APIs, and framework integrations connect your application to VeriProof. Developers can start with the SDK quickstart. Once signed in, they'll also find SDK details under Tools and SDKs in the customer portal. Explore integrations →

Five: Allow (continue as submitted), Warn (continue with a warning), Transform (continue with a corrected version), Escalate (wait for a person with authority), or Deny (stop). Your application acts on the result. Review policy decisions →

A policy can route an action to an authorized person when judgment or approval is required. VeriProof keeps the context, reviewer decision, reason, and related evidence connected to the workflow record. Explore human review →

The action doesn't run. The SDK runs the protected action inside the check, so the action starts only after an allowed result. A timeout or failure comes back as that status, with a reason. Your application then decides whether to retry or hand the work to a person. Explore policy decisions →

It depends on the change. A new model, tool, or permission can alter how a workflow behaves. Your team's change process determines the testing and approval needed.

Your organization determines which requirements apply, with advice from its specialists. VeriProof connects those requirements to controls and evidence for review. Explore regulatory compliance →

Teams can see which controls and evidence support a requirement, then assign work where something is missing. Findings and follow-up stay connected to the applications involved.

An evidence pack brings the review scope, policy decisions, human review, application outcomes, and supporting records together. Its contents depend on the engagement and the evidence your team chooses. See an evidence pack →

Yes. The sample pack is the one a reviewer would receive for a supplier-invoice agent: one record per decision, the redaction report, and the integrity proof. Download it and check it with your own risk or audit colleagues. Download the sample evidence pack →

Each environment uses one proof provider, fixed when the environment is created. Azure SQL Database Ledger is the default. A SaaS environment can use the Solana blockchain instead, which receives only fingerprints, never personal or patient data. Private Enterprise uses Azure SQL Database Ledger. VeriProof's main database is never used as the proof. How record integrity works →

VeriProof provides the evidence for the assessment. Your organization and its appointed reviewers draw the conclusion from the full assessment.

Gaps become work your team can track. The Compliance workspace connects each gap to affected applications, an owner, and a deadline, with context available for audit review.

Choose VeriProof-managed SaaS today. Private Enterprise is coming soon to Azure Marketplace. It will run entirely in your own Azure subscription, completely separate from veriproof.app, integrity ledger included. Explore Private Enterprise →

The SDKs and the platform use the standard VeriProof software license; they are not open source. Source is available on request. On the SaaS plans, including the trial, you get the security review tier: you can read and inspect the code. The modification tier lets you change the code and run it in your own environment. It comes only with Private Enterprise and is set in your agreement. Releases are signed and reproducible, so you can confirm the code you reviewed is the code you run. How source access works →

That depends on the policy. Your team selects the facts needed to check an action and the records worth retaining for review. Explore data safeguards →

Your administrator creates and controls the application credential used by the connected workflow. Do not send credentials through public forms or evaluation requests.

The public security pages and Trust Center cover the service's controls and data handling. If your review needs more detail, contact VeriProof about the relevant materials. Explore security →