High-Velocity AI Context
Engines for LSPs
The Smarter Way to Pre-Translate Your Continuous Volumes.
Stop wasting your internal team's time on generic, out-of-context machine translation that takes forever to clean up. Hourglass delivers enterprise-grade, hyper-contextualized AI pre-translation engineered specifically for the demands of modern Language Service Providers.
We process your continuous file streams through advanced AI pipelines that adapt to your client's exact terminology and style before your human editors even open the file. You get flawless, CAT-tool-ready bilingual files, plus a ready-to-use production environment for your linguists.
Connect Your PipelineBuilt for Production at Scale
Built for Continuous Volumes, Not One-Offs
We deliver maximum efficiency on your regular, ongoing account workflows. Continuous cooperation allows our engines to ingest your historical data — meaning the AI's precision and terminology match compound over time.
Absolute Data Privacy & Security
Full GDPR compliance. Your data is strictly confidential and never used to train public models. Every single word is processed exclusively on secure, high-performance EU-based servers.
Maximum Digital Leverage
Just 2 cents per processed word. Inject a premium, fine-tuned AI workflow directly into your production line for a fraction of traditional engineering costs.
Smart Filtering. Protecting Your TM and Your Budget.
Hourglass is fully compatible with modern CAT tool logic, safeguarding your data integrity and preventing unnecessary costs at every step.
Locked Segment Exclusion
Our engine automatically detects and excludes all locked segments from processing and billing. This protects your translation memory integrity and prevents any unnecessary costs.
Bilingual Delivery with Quality Comments
You receive a fully prepared XLIFF file containing the translation alongside AI-generated comments detailing specific adjustments and stylistic adaptations. Your post-editors work twice as fast.
Free Post-Editing Environment
If your external linguists lack their own CAT tool licenses, we provide a secure, browser-based post-editing workspace at no extra cost. Log in, upload the file, view AI comments and glossaries, complete the task — zero software overhead.
Strict Scope Filter
To guarantee maximum contextual accuracy, we only process complete, continuous text. We do not accept or translate isolated word lists, software strings, standalone glossaries, or fragmented texts lacking proper context.
We Adapt to Your Ecosystem Instantly
Send us your volume files in your preferred formats and receive them back ready for your post-editors. No workflow restructuring required.
Supported Document Formats
Native CAT-Tool Compatibility
What We Need to Start
To ensure absolute terminology match and style alignment from the very first batch, we build a dedicated contextual profile for your recurring accounts. To onboard a regular stream, we require:
-
01
Source Files
The source files from the specific project requiring processing, in any of our supported document or CAT-tool formats.
-
02
Style Guides & Reference Materials
Your client's specific style guides, glossaries, or reference materials. The more context we have, the more precisely our engine adapts.
-
03
The Blueprint Pack
Provide 5 to 10 previously translated, edited, and fully approved bilingual files from the specific project. These finalized files are critical to train the context engine for maximum quality — this is the most valuable input we receive.
Most Important
Context Compounds Over Time
Each batch we process refines the AI's understanding of your client's voice. The more volume you run through Hourglass, the sharper and faster the output becomes.
Your First 1,000 Words
Are on Us
Ready to see how much faster your team can work with hyper-contextualized AI files? Submit your project onboarding files along with your Blueprint Pack, and we will process your first test file of up to 1,000 words completely free of charge.