Post by Phrase

Do more with less. It's a mandate most teams know well, yet few have approached it as thoughtfully as the Marketing Localization team at Wellhub. In a fantastic article for MultiLingual Media, Josevi Abad, Localization Director at Wellhub, shares how the team turned years of linguistic investment into a foundation for smarter automation. They already had something incredibly valuable. A repository of translation memories, carefully built and maintained by their in-house linguists. Using Phrase Custom AI, they trained neural engines on that data while retaining full control of their linguistic assets. As Josevi puts it: "This exploration led us to our solution: Phrase's Custom AI. This choice represented a fundamental divergence from the GenAI-for-everything trend. Unlike generic models that act as black boxes trained on the entire Internet, this technology trains neural machine translation (NMT) engines exclusively on our own linguistic assets." This story highlights a few ideas that many teams will recognize: • Your existing linguistic assets may be more valuable than you realize • The quality of AI output is shaped by the quality of the data behind it • Automation creates more space for creative and cultural expertise A big thank you to Josevi and the wider Wellhub team for sharing their experience so openly. We loved reading this and think many teams facing similar challenges will too. Link to the full article in the comments.