At Unbabel we truly believe that you can only “solve” translation by augmenting humans with artificial intelligence.
First, our domain-adapted machine translation already out-performs general purpose MT from Google, Microsoft and others, which we then enhance further with custom glossaries and style guides on a per customer basis.
Only then does it get distributed to our global network of 100,000 Unbabelers, who bring everything up to human quality by smoothing out hard-to-catch errors and ensuring accurate and fluent translations.
What guarantees quality at Unbabel is not the quality of the individual translators themselves — which even at a professional level can vary widely and are naturally prone to human error — but the quality of the pipeline that produces the work in precise, uniform and accurate steps.
Human work is still required, but at non-critical junctures where they correct and edit machine work, rather than tasked with the entire job — we leave the grunt work to our award-winning artificial intelligence engines and our community can focus on what they’re really good at. This removes human dependency, but it vastly increases their value for quality corrections overall, and allows for an exponentially higher throughput of content.
For you, this means that high volumes of customer emails, chat, FAQs and other dynamic forms of content can be translated in parallel across 28 languages in near real-time while maintaining a consistent quality and being cost-effective. Neat, right?
If you want to know more about how we scale translation quality and speed, give this a read.
Speed is of no essence....!
Please sign in to leave a comment.