
Craig Stewart
Craig Stewart is the Director of AI Research at Phrase. A graduate of Carnegie Mellon, he is an expert in AI with a particular focus on MT Evaluation. Notably, he was one of the primary architects of the COMET metric, which is the current industry standard for measuring MT quality. Craig also has a background in translation and linguistics and is passionate about designing AI solutions that benefit the human experience.
Editorial Articles
-

When AI Scales Content, What Defines Quality?
Craig Stewart argues that as AI scales content globally, quality must be redefined by business impact, not just linguistic accuracy.
-

Resonance Over Volume: The Future of Multilingual Content
Eddie Arrieta and Craig Stewart uncover why accuracy isn’t enough anymore and how AI must prove impact, not output, to drive real global growth.
-

What If Everything You’re Measuring Is Missing the Point?
Traditional translation metrics miss user impact. Craig Stewart urges outcome-based evaluation to track real-world engagement and results.
-

Reengineering Localization To Drive Engagement
Stean Huyghe explores how AI is reshaping localization, shifting from translation to strategy, in a discussion with Craig Stewart, Director of AI Research at Phrase.
-

Reengineering Localization for the GenAI Age
Stefan Huyghe and Craig Stewart discuss how GenAI is transforming localization, from MQM limits to new roles for linguists in global content strategy.
Blog Posts
-

Stop measuring accuracy. Start measuring impact.
Generative AI has changed the economics of content. Businesses can now produce material at unprecedented scale and speed. For leaders under pressure to expand globally and accelerate growth, this looks…
-

Phrase Quality Performance Score: What You Need to Know
Phrase Quality Performance Score streamlines and scales translation workflows, optimizing costs and time for efficient localization.
Webinars
-

Introducing Phrase Quality Performance Scores—the next step in AI-powered quality assessment
Join us for an exclusive unveiling of Phrase Quality Performance Scores — our new and improved automated method for calculating quality scores, based on the latest MQM 2.0 framework!
-

Revolutionize Your Language Assets: Introducing Automated Asset Curation
Join us for a webinar where we unveil the future of language asset management with Automated Asset Curation (AAC). Discover how this revolutionary capability automates the cleaning and curation of…
-

Phrase Bootcamp: Uncover the Full Potential of the Phrase Platform
The Phrase Bootcamp is your opportunity to explore the full capabilities of the Phrase Language Intelligence Platform—tools designed to optimize your workflows and achieve tangible results.