Engine 04 · Available now

Stop Data Leaving
Without Sending It Away.

Data loss prevention inspects your most sensitive content by design. nPro DLP does that inspection on your own infrastructure.

Most security tooling watches the wrong direction

Detection is built to keep attackers out. Data loss prevention works the other way: it watches what leaves, and stops sensitive information reaching places it should not go.

That covers accidents as much as attacks. A customer database attached to the wrong email. Credentials pasted into a public tool. A departing employee copying files to personal storage. None of these are a breach in the traditional sense, and none are caught by tooling aimed at intrusion.

For regulated organisations the distinction is not academic. Under GDPR, PDPA and similar frameworks, an accidental disclosure carries the same reporting obligation as a deliberate one.

Two ways to inspect. Only one keeps the data.

Cloud DLP requires you to send the vendor exactly the content you are trying to protect. That is a difficult position to defend to an auditor.

Cloud DLP

Your network

Customer records

Vendor cloud

Inspection happens here

The protected content crosses the boundary in order to be protected.

nPro DLP

Your network

Customer records
Inspection happens here

Nothing crosses the boundary. Residency becomes a property of the architecture rather than a clause in a contract.

Same stack

Runs alongside XDR, SIEM and EMS, so egress events land on the same timeline as everything else.

Same deployment

No separate product to procure, install or learn.

Same pricing basis

Flat per agent, with no charge based on how much content is inspected.

Coverage

Eight ways data leaves an endpoint

Each one is inspected on the machine itself. The content never leaves your estate in order to be checked.

USB and removable devices

Files written to USB drives and other removable media.

Sensitive data discovery

Scans endpoints for regulated content sitting at rest, so you know where it is before it moves.

Clipboard

Sensitive content copied and pasted between applications, recorded against the real destination, so a paste into Gemini shows as Gemini.

Print

Sensitive material sent to local and network printers.

Cloud uploads

Uploads to personal cloud storage and other web destinations, including text inside screenshots and photos.

Messaging

Files and text sent through messaging applications.

Screen capture

Screenshots and screen recording of sensitive content.

Email

Outbound email is flagged through your mail platform’s audit feed. Inline gateway blocking is the next phase, not something the engine does today.

Monitor today

Rollout

Nothing blocks until you say so

DLP is the one control that can stop people working. So it ships off, and every step towards enforcement is deliberate.

01 · Disabled

Off by default

The engine ships disabled. Nothing is inspected until you turn it on.

02 · Monitor

Watch first

Events are recorded and reported, nothing is stopped. You see what your policy would have caught before it catches anyone.

03 · Audit

Warn, then allow

The employee sees why the action is risky and can carry on. The event is recorded, so you see how often a rule would fire before it blocks anyone. Warnings can be shown in English, Bahasa Malaysia or Bahasa Indonesia.

04 · Enforce

Then act

The action is blocked. Switching a group into enforce means typing a confirmation phrase, so no fleet goes into blocking by accident.

Per group, not per fleet

Posture is set per group, so finance can sit in enforce while everyone else is still in monitor or audit.

Policy survives the network

Agents cache the last policy they were given, so a laptop off the network keeps the posture you set for it.

Start from a policy, not a blank page

Nine ready-made policy templates cover the common starting points. Before one is applied, you see exactly what will change.

PDPA baseline Malaysian banking Indonesian banking Healthcare Legal contracts HR records Source code Strict AI agent General enterprise

Rules that say what you mean

Build rules from how sensitive the data is, how risky the destination is, which channel it is going through and the time of day. For example: block restricted data to any high-risk AI tool. If a rule can never trigger, you are told before you save it.

Detection

What it recognises

Patterns for the identifiers you are obliged to protect, a model for the documents that have none, and your own rules for everything specific to you.

Southeast Asian national IDs

Malaysian NRIC, Indonesian KTP and NPWP, Thai national ID, Singapore NRIC and FIN, Vietnamese CCCD and Philippine PhilSys. Checksums and nearby keywords keep an ordinary 12-digit number from being mistaken for an ID.

ML classification

A machine-learning model recognises sensitive documents by what they mean, such as contracts, financial reports and strategy papers, even when they contain no ID or card numbers at all.

Text inside images

Screenshots and photos uploaded to the web are read for text and checked like any other document, so a photo of an IC, a bank card or a payslip no longer passes as just an image.

Your own detectors

Build patterns for internal project codes, customer or policy numbers in the Control Panel, and test them on sample text before they go live. No waiting on the vendor.

Evidence

What you get once it is on

PDPA compliance reports

A ready-made report on personal data found, cross-border transfers, blocks and unapproved AI use. PDF or CSV, emailed automatically daily, weekly or monthly, for the people who ask for evidence rather than dashboards.

MITRE ATLAS and ATT&CK tagging

Every DLP alert is labelled with the framework technique it represents, so security teams read it in the same language as the rest of their SIEM.

Graded alerts in Npro SIEM

Alerts are ranked by severity, so the serious ones surface first, and land on the same timeline as detections from XDR rather than in a console of their own.

12 months of history

DLP events are kept for 12 months, on your own servers, so a year of evidence is there for audits and investigations.

Fleet coverage

One view of which endpoints are running the engine and which are not, so gaps are visible rather than assumed.

Tamper protection

Employees without admin rights cannot stop or uninstall the nPro agent, and if it crashes it restarts itself. A DLP agent that employees can switch off is not protection.

Try it for 30 days, free

Leave DLP in monitor and look at what a month of real traffic tells you. Nothing is blocked, nothing leaves your network, and there is no charge based on how much content is inspected.

Or compare nPro DLP with 15 other platforms →