Case Study: Ariel Corporation achieves major translation-quality gains and cost savings with SYSTRAN's Neural Fuzzy Adaptation

How to harness extra value from your language assets

Ariel Corporation needed to get more value from its language assets—reducing post‑editing time and costs while preserving translation quality and brand voice. To do this they adopted SYSTRAN machine translation integrated into XTM’s AI‑enhanced Translation Memories (the neural fuzzy adaptation in XTM 12.7), with their CMS feeding XTM automatically to streamline localization and customization.

SYSTRAN’s NMT was configured to use validated fuzzy matches from Ariel’s translation memories as real‑time references, boosting MT output and reducing human effort. The combined Neural Fuzzy + Specialization approach delivered an expected +22 BLEU‑point improvement (18 + 4), lowered post‑editing and costs, increased productivity, and produced more consistent, higher‑quality translations—helping Ariel harness both machine translation and human editing for outstanding cost efficiency.

Ariel Corporation

Ronald Egle

Content Systems Administrator