What tools do companies use to connect localization to product workflows?

The localization-product workflow gap

Most localization problems in product organizations are not translation quality problems. They are workflow problems. Strings are added to the product, manually exported to a spreadsheet or file, emailed to a localization team or vendor, translated, reformatted, and re-imported — a cycle that takes days or weeks and requires manual effort at every step. By the time translated strings are ready, the product has already shipped in English, and localized versions fall further and further behind.

The solution is not to translate faster. It is to remove the manual steps that create the gap. When localization is connected directly to the product workflow, new strings are detected automatically, translation jobs are created without manual effort, and translated content is delivered back to the repository or design tool without requiring anyone to manage the handoff. The localization cycle runs in parallel with the product development cycle, not after it.

The tools that make this possible fall into three categories: repository connectors, design tool integrations, and direct API integration.

The three integration approaches for connecting localization to product workflows

1. Repository connectors (GitHub, GitLab)

Repository connectors bridge your code repository and your translation management system directly. When developers commit new or updated resource files, the connector automatically detects the changes, uploads the new strings to the TMS, and triggers the configured translation workflow. When translations are complete, the connector creates a pull request with the translated files, allowing the localization update to merge into the codebase through the same review process as any other code change.

This approach is ideal for software products and mobile applications where strings are stored in resource files in the codebase. It eliminates the manual file export and import cycle entirely and enables engineering teams to include localization status as part of their standard CI/CD checks, blocking merges until all strings are translated and approved.

Smartling's Repository Connector supports GitHub and GitLab, automatically scanning repository resource files for new content, creating localization branches, and delivering translated files back through the pull request workflow. The connector is designed for continuous deployment environments where localization speed is as important as localization quality.

2. Design tool integrations (Figma)

Design tool integrations connect the localization workflow to the design phase of the product lifecycle, where strings often originate. Rather than waiting for strings to be hardcoded before sending them for translation, design integrations allow teams to begin localization during the design review stage, when changes are still low-cost to make.

Smartling's Figma plugin allows designers to upload design files directly to Smartling for localization, enabling translated strings to be reviewed in the context of the design before any code is written. This catches layout issues, character expansion problems, and cultural concerns at the design stage, where they are cheapest to fix.

3. Translation management system API

The TMS API gives engineering teams direct programmatic access to all platform capabilities: uploading strings, creating jobs, triggering workflows, checking translation status, and downloading completed translations. This approach requires development effort to implement but provides the most flexibility for teams with custom content pipelines, proprietary content management systems, or specific workflow requirements that do not map to a pre-built connector.

API integration is also used for embedding localization into CI/CD pipelines: automated build systems can query the TMS API to check whether all strings for a release are translated and approved before allowing a deployment to proceed.

Key tools for connecting localization to product workflows

The specific tools that product and engineering teams use most commonly for localization integration fall into four categories.

Repository connectors

GitHub and GitLab connectors are the most common integration point for software product teams. Smartling's Repository Connector monitors the configured repository for changes to resource files, automatically uploading new strings to the TMS and delivering translations back as pull requests. The connector supports multiple file formats and can be configured to handle different content types with different workflow rules within the same repository.

Design tool plugins

Figma is the dominant design tool for enterprise product teams, and Smartling's Figma plugin integrates localization directly into the Figma workflow. Designers can upload design files to Smartling for translation without leaving Figma, and translated strings can be reviewed in the context of the original design layout. This enables earlier localization review cycles and catches design-level localization issues before they become engineering problems.

Translation management system API and SDKs

Smartling's RESTful API provides access to the full platform capability set for engineering teams building custom integrations or embedding localization into automated build and deployment workflows. Software development kits (SDKs) are available to reduce the development effort required to integrate Smartling API functions into existing code. The API is also used for CI/CD integration, where build systems check translation completion status before allowing deployments to proceed.

Project management and collaboration integrations

Some engineering teams use localization workflow integrations with project management tools to track translation status alongside other development tasks. Smartling integrates with tools in the product development ecosystem to provide visibility into localization status within the workflows that product and engineering teams already use for project tracking.

How continuous localization through product workflow integration works

Here is how a continuous localization workflow runs when localization is connected directly to the product development cycle:

When product workflow integration may not be the immediate priority

Enterprise checklist for evaluating product workflow localization integration

Use these questions to assess whether a translation management platform can integrate effectively with your product development workflow.

Repository connector

• Does the platform offer a certified repository connector for your code repository platform, specifically GitHub or GitLab?
• Does the connector monitor the repository automatically for changes to resource files, or does it require manual triggers to initiate content upload?
• Does the connector deliver translations back to the repository as pull requests, enabling translation merges to go through the standard code review process?
• Can CI/CD integration be configured so that builds check translation completion status before allowing deployments to proceed?

Design tool integration

• Does the platform offer a Figma plugin or equivalent design tool integration for your team's primary design environment?
• Does the design tool integration support bidirectional sync: uploading strings from design files and delivering translated strings back for in-context review?
• Can translated strings be reviewed in the context of the original design layout within the design tool, enabling layout and character expansion issues to be caught before engineering handoff?

API and SDK

• Does the platform offer a RESTful API with full access to platform capabilities, including job creation, workflow triggering, status checking, and translation download?
• Are software development kits (SDKs) available for your team's primary development languages to reduce the effort of API integration?
• Is the API designed for use in automated build systems, with appropriate rate limits, authentication, and status endpoint design for CI/CD use cases?

Workflow configuration and automation

• Can different string types in the same repository be routed to different translation workflows automatically, based on file type, path, or metadata?
• Does the platform support job automation rules that batch strings and create translation jobs automatically without manual intervention?
• How is translation memory and glossary applied for product strings: from the first-pass AI output, or only during human review?

How Smartling connects localization to product workflows

Smartling provides three integration paths for connecting localization to product development workflows, each designed for a different point in the product lifecycle.

The Repository Connector bridges GitHub and GitLab repositories directly to Smartling. When developers commit new or updated resource files, the connector automatically detects the changes, uploads strings to Smartling, and triggers the configured translation workflow. Completed translations are delivered back as pull requests, enabling the localization merge to go through the standard code review process. The connector is designed for continuous deployment environments, with support for CI/CD checks that verify translation completion status before deployments proceed.

The Smartling Figma plugin allows designers to upload design files to Smartling directly from Figma, enabling translated strings to be reviewed in the context of the original design layout before engineering handoff. This moves localization review earlier in the product cycle, when design changes are still low-cost to make.

Smartling's RESTful API provides full programmatic access to platform capabilities for teams building custom integrations or embedding localization into automated build and deployment systems. SDKs are available to reduce development effort. The API supports CI/CD integration patterns including translation status checking as a build gate.

Across all integration paths, Smartling's AI Adaptive Translation Memory, glossary enforcement, and AIHT workflow ensure that product strings are translated with the same quality standards as other content types. Job automation rules batch and route strings automatically, and approved translations are written back to translation memory to continuously improve future AI output for similar product content.

Smartling is rated the number one enterprise translation management system on G2 for 20 consecutive quarters, and holds ISO 27001, SOC 2, HIPAA, HITRUST e1, PCI Level 1, and ISO/IEC 42001:2023 certifications.

Related questions