We are expanding DeepL’s data infrastructure to deliver the global scale and performance our customers expect—and to ensure that as our user base grows across every continent, the quality of every translation experience improves rather than degrades.
Why Expansion Is Necessary
As DeepL adoption has accelerated worldwide, our user base has expanded to more than 100 countries and regions. Users and enterprises across the globe access DeepL in different time zones and under vastly different network conditions. To deliver the optimal experience, our infrastructure needs to move closer to users, minimizing latency and ensuring service availability regardless of geographic location or local internet infrastructure quality.
The physics of data transmission are unforgiving: light travels approximately 200 kilometers per millisecond through fiber, and every network hop adds processing overhead. A user in Singapore accessing servers in Frankfurt experiences a minimum of 150ms of round-trip latency before any processing even begins. For real-time applications like Voice-to-Voice translation, where the entire pipeline must complete in under 200ms, this delay alone would make the service unusable. Expanding infrastructure to serve users from points of presence in their region is not a luxury—it is a mathematical necessity.
Adding AWS as a Sub-Processor
We are adding Amazon Web Services (AWS) as a sub-processor to our data infrastructure. This enables us to deploy service nodes across multiple global regions, reducing translation service latency to under 100 milliseconds regardless of where users are located. AWS’s global infrastructure footprint, spanning more than 30 geographic regions and over 100 availability zones, provides the coverage necessary to bring DeepL’s AI models close to every major user concentration on the planet.
This multi-cloud architecture, combining our existing infrastructure with AWS’s global reach, creates redundancy and resilience that a single-provider approach cannot match. Service continuity is maintained even if a regional outage affects one provider, and load can be dynamically balanced across regions based on real-time demand patterns. For enterprise customers with SLA commitments, this architectural resilience translates directly into higher uptime and more predictable performance.
Privacy and Compliance
While expanding our infrastructure footprint, we remain steadfastly committed to maintaining the most rigorous global privacy standards. Every aspect of our data handling architecture has been designed to meet or exceed the requirements of the most demanding enterprise compliance frameworks:
- All data is encrypted with AES-256 both in transit and at rest, ensuring that even if network traffic were intercepted, the content would remain unreadable
- DeepL Pro user data is never used to train AI models, a contractual and architectural commitment that distinguishes DeepL from consumer-grade translation services where user data feeds model improvement
- Data is deleted immediately after translation completion, with no persistent storage of translated content
- Full compliance with GDPR and SOC 2 Type II standards, independently audited and continuously maintained
- Customers can select data processing regions (EU/US) to meet jurisdictional data residency requirements
AWS’s infrastructure is certified under ISO 27001, SOC 1/2/3, PCI DSS, and other global security standards, fully aligning with DeepL’s own security requirements. This isn’t a compromise between performance and compliance—it’s a demonstration that the right architectural choices can deliver both simultaneously.
What This Means for Users
For most users, this change is seamless and transparent. You may notice improved translation response speeds, particularly outside of Europe where the infrastructure expansion has the most measurable impact. For enterprise customers who require additional detail on data processing or have specific data residency requirements, please contact your customer success manager or sales team for a detailed briefing on how the expanded infrastructure affects your specific deployment.
Our commitment remains unchanged: to deliver the world’s most accurate and secure AI translation service while continuously improving performance and reliability. Infrastructure expansion is a quiet but essential component of that commitment—invisible when it works perfectly, but the foundation upon which every millisecond of performance and every percentage point of uptime depends.