AI assistants that help you stay on budget: a case study

When budgets are tight and quality can’t be compromised, innovation is key. 

The challenge:

A client in the financial-technical services domain regularly ordered machine translation with single-step human post-editing as a cost-saving measure. Adding both a translator and an editor was not possible due to budget constraints. At the same time, the client expected top-quality translations. But the results consistently fell short.

Languages: English → Latvian, Estonian, Lithuanian.

Reference materials: translation memory, glossary, and client’s instructions.

The solution:

Our head of localization, Maksym Hamii, developed custom AI-powered assistants specifically tailored to the client’s projects.

We trained the assistants on the client’s language pairs and project reference materials, then integrated them into a new AI-powered localization workflow:

Machine translation → AI-powered revision → Human post-editing

The result:

• Edit distance increased to 92.6%, showing significantly higher translation quality
• Human-sounding, contextually accurate output
• Projects kept going within the same budget and timeline, but with much better results

This success story, along with the broader AI revolution in localization, inspired us to develop Hera AI — InText’s own localization tool that amplifies quality without increasing costs. Stay tuned to our upcoming posts to learn more. 

Let’s make your next localization project both budget-friendly and high-quality!