How Nord Security Automates Localization Under Strict Guardrails

Nord Security is a global cybersecurity leader, trusted by over 20 million users worldwide. As the team behind great names like NordVPN, NordPass, and NordLocker, they operate on the absolute cutting edge of digital protection.

Mantas Baraišis, Head of Localization at Nord Security, told us what it takes to localize products in a highly secure industry where brand communication requires maximum compliance.

To manage massive, synchronized launches across their massive software ecosystem, Mantas and his team built an infrastructure with zero data leaks and maximum efficiency.

Here is how they did it.

Why the key to continuous localization starts with engineering

Before Crowdin, localization at Nord Security was managed via Google Sheets and Excel files. For a company growing at a breakneck pace, manual file sharing was a ticking clock of human error, slow time-to-market, and broken formats.

Mantas knew that to truly scale, the localization team (which sits inside the marketing department) had to remove itself from the technical pipeline entirely. He went straight to the engineering teams with a simple pitch:

"

Invest just two weeks of your time right now to set up the automation, and we will only ever need to talk during team-building events or corporate parties.

Developers, naturally focused on optimization and reducing long-term maintenance baggage, immediately agreed.

Infrastructure

Using Crowdin’s Console Client (CLI) tool, the developers integrated Nord Security’s code repositories directly into the localization workflow. Today, whenever engineers push new features, the strings (JSON, XML, or other formats) are autonomously pulled into Crowdin, put through the localization workflows, and pushed back without a single manual intervention.

Nord Security has been building this flawless, continuous infrastructure with us since 2022, leveraging the advanced capabilities of Crowdin Enterprise. The platform’s architecture allowed the company to split complex workflows between internal marketing teams, developers, and external vendors while maintaining absolute data security.

100% human quality: Protecting the brand in a high-stakes market

In the cybersecurity industry, cost-cutting in translation can lead to reputational problems. While many localization teams are rushing blindly into raw generative AI translations, Nord Security enforces a strict 100% human-in-the-loop review policy.

Operating in a high-stakes sector, Mantas emphasizes that a single unnatural translation or an awkward localized button can trigger user suspicion, closely resembling a fraudulent phishing scam. For a global brand built entirely on digital protection and data integrity, compromise is not an option.

Balancing human expertise and internal AI guardrails

To handle a massive product ecosystem spanning web, mobile, and browser extensions without burning out their team, Nord Security cleverly divides its resources:

• Focus markets – 13 core regions are strictly managed by internal copywriters who possess deep contextual knowledge of the product ecosystem.
• Secondary markets – translated by trusted external vendors. However, to maintain total consistency, internal copywriters perform a mandatory second-pass proofreading after the agency.

"

Being the industry leader, we spare no effort to have the best quality. And it goes with any part of the company. Depending on the company cycle, you can usually pick two: Growth, Operations, or Cost-cutting. We are Growth and operations-oriented at the moment in time.

When it comes to technology, Nord Security actually built and launched its own enterprise AI optimization platform, Nexos.ai. Yet, due to their rigorous data training privacy, and strict risk compliance policies, the company maintains a highly cautious approach to third-party generative AI tools.

Instead of relying on AI-generated translations, they optimize their workflow by augmenting human translators with translation APIs such as DeepL and Lara. From a manager’s perspective, these engines are updated with the latest translation memories and term bases directly through the Crowdin Enterprise interface – ensuring maximum speed and consistency without a single security leak.

"

Both engines cover languages differently from one another, and it’s a matter of preference of the translator/copywriter which is better for their language. From a manager’s perspective, both of these MT engines get easily updated with the latest TMs and TBs from the Crowdin interface, which makes our lives very easy.

Visual context challenge for mobile UI

Mobile app interfaces are notoriously where localization goes to die. Without visual context, strings overlap, translations break layout constraints, and buttons break.

To ensure their localized apps maintain a consistent look across devices, Nord Security integrated manual and automated context loops directly into their pipeline. Whenever developers push new strings to Crowdin Enterprise via the CLI, they pass direct Jira or Figma links into the context fields. This gives internal copywriters immediate visual references for exactly how much spatial layout they have to work with.

"

With some teams, we have a setup that whenever they are pushing new strings to Crowdin, they add a Jira/Figma link in the context field. Some do upload screenshots and map them. And with some, we have a bot that takes translation screenshots of the release and posts them regularly in a Slack channel.

For secondary markets and continuous releases, they even automated the process further: a custom bot regularly takes translation screenshots of new builds and drops them into a dedicated internal Slack channel. That is how Nord Security ensures the mobile UI looks perfect before any release goes live.

Correlation between analytics and device locales

Many companies evaluate localization success purely by tactical output – like the volume of translated words. For Nord Security, the strategy is strictly tied to real business growth and performance data.

Mantas and his team actively track specific performance signals to guide their localization investments:

"

Main metrics we track are device language: per country, revenue split vs EN, locale CR [conversion rate] vs EN. If any of these are tipping the scales more towards the localized version, then we consider investing more time. And of course, if the language maintenance cost does not exceed the revenue generated by it.

For example, they recently noticed a performance pickup in a country within Southern Europe. However, data analysis showed that less than 30% of locals set their laptops and iPhones to the native language. Because the captured market size was too small, the team determined that localizing the entire app wouldn’t provide a strong ROI, whereas localizing the web interface made much more financial sense.

By keeping localization deeply integrated into core marketing data, the team proves that localization actually drives revenue.

Summary: Why Nord Security chooses Crowdin Enterprise

Reflecting on Nord Security’s localization journey, Mantas shared the specific features and advantages that make Crowdin Enterprise their long-term platform of choice:

• Developer-first simplicity: Mantas loves how easily engineers can adopt the platform. He highlighted that the straightforward UI and clear documentation allowed their developers to configure repository automations via the CLI tool in no time.
• Easy asset maintenance: From a management perspective, Mantas highly values how simple it is to keep translation assets current. He loves that updating machine translation engines with the company’s latest corporate glossaries and translation memories takes just a quick click right inside the interface.
• Flexibility to build custom extensions: Mantas appreciates that Crowdin’s adaptable ecosystem allowed his team to develop their own plugin.