Case Study

11 days, 70% improvement:

How SNOMED International proved that AI agents can translate industry specific terms with speed, precision and accuracy

SNOMED International is the developer of the world’s leading clinical terminology, a “dictionary” of precise medical terms that enables collaboration between healthcare systems across the globe.

The terminology has been well received in the anglosphere, but increasing adoption requires translating into multiple languages; each translation is currently an expensive, manual and time-consuming process.

In just 11 days of working with Veratai, SNOMED proved that an AI agent could accurately translate precise clinical terms into native languages (even languages that are low on reference material), paving the way for SNOMED to increase adoption worldwide.

Before:

expensive manual translation

  • Translation required a large specialist team of bilingual, clinically trained specialist.

  • Full translation of the terminology cost in excess of €1 million.

  • Existing automated services (Google Translate, DeepL) produced a low-quality translation, requiring significant human intervention to use.

  • Rolling out the terminology to new languages was a hard sell to member countries, who struggle to justify a large-scale translation programme.

After:

proof that automatic translation is viable

  • 76% of translations using Veratai’s AI agent met the standard for clinical usage, up from 44% using an off-the-shelf service

Bespoke Agent Design

Our Method for helping SNOMED achieve results

  1. Start with the human process: Veratai analysed and understood the current process followed by human translators.

  2. Discover the right data: We prepared training and test data, including:

    • Paired translations

    • Reference materials in the target language

    • A translation style guide

    • Relevant academic works

    • A corpus of relevant medical texts in the target language, which could be cross-checked to verify the translations

  3. Develop the agent: We built a bespoke agent, giving the LLM flexibility to select from different translation resources and iterate on the translations, in just 3 days.

  4. Evaluate the results: With TEHIK, the Health and Welfare Information Systems Centre in Estonia as a partner, we “blind evaluated” translations of 100 pieces of vocabulary from the terminology and measured the results

From confusion to confidence

By performing a quick development and evaluation phase of an AI agent, SNOMED were able to confidently present the translation results at their annual conference