Log inRequest a demoStart free

From discovery to a certified QA dataset

Your network does the data work. DataNivra Cloud coordinates.

Eight steps. Each one is labelled with where it runs.

Eight steps, each in the place it runsSteps 1 to 3 (install the agent, register sources, discover) run in your network. Steps 4 and 5 (approve policy, request a dataset) happen in DataNivra Cloud. Steps 6 and 7 (build and certify, provision) run in your network again, and step 8 (track status and evidence) is in DataNivra Cloud. Only metadata passes between the two.DataNivra Cloudpeople and policyYour networkrows, keys and processing1Install2Register3Discover4Approve5Request6Build7Provision8Track

  1. Install the agent

    Runs in your environment

    Run the containerised agent inside your VPC, VNet or data centre. It enrols with a one-time token and from then on connects outbound only.

  2. Register sources by reference

    Runs in your environment

    Point the agent at databases, lakes or files using secret references (for example a Key Vault or Vault path). Credentials are resolved locally and never sent to us.

  3. Discover and classify

    Runs in your environment

    The agent reads schema metadata and profiles locally, then reports sensitive-field findings as column names, classes and counts.

  4. Define policy and approve

    DataNivra Cloud · metadata only

    Data owners review findings and approve versioned classification, masking and subset policies in the console, with separation of duties.

  5. Request a dataset

    DataNivra Cloud · metadata only

    A tester or pipeline requests a dataset: business entity, scope, masking policy, target environment, refresh and retention.

  6. Build, validate and certify

    Runs in your environment

    The agent leases the job, re-verifies the policy checksum, subsets, masks or synthesizes, validates referential integrity and runs certification gates.

  7. Provision and refresh

    Runs in your environment

    Only certified datasets are provisioned to the target environment. Refreshes rebuild and re-certify on schedule.

  8. Track status and evidence

    DataNivra Cloud · metadata only

    The console shows job state, aggregate metrics and evidence references (URI + checksum). The evidence files stay with you.

A developer working with code on a laptop and a second monitor

Testers and pipelines request datasets; the agent in your network builds them.

Outbound lease modelThe control plane queues a declarative command and grants a time-bound lease when the agent asks for work. The agent validates the command and the policy checksum, executes locally and reports metadata. The control plane never opens a connection into your network.DataNivra CloudQueue declarativecommandGrant time-bound leaseRecord status & auditYour environmentAgent polls for workVerify command &policy checksumExecute locallyReport metadataAgent-initiated HTTPS only; no inbound ports
Outbound lease model. The control plane queues a declarative command and grants a time-bound lease when the agent asks for work. The agent validates the command and the policy checksum, executes locally and reports metadata. The control plane never opens a connection into your network.
Text description
  1. DataNivra Cloud: Queue declarative command → Grant time-bound lease → Record status & audit

    Connection: Agent-initiated HTTPS only; no inbound ports

  2. Your environment: Agent polls for work → Verify command & policy checksum → Execute locally → Report metadata
Start freeTry the sandboxSee how it worksInstall agentView docs

Prefer a conversation? Request a demo or contact sales — optional, never required.