We've been exploring the use of Smartling's service as a key part of introducing machine translation into the localization of SuMO's KB articles, but the lack of a straightforward and programmatic way to update Smartling's translation memory with the feedback/adjustments of our localization leaders and contributors is a deal-breaker. Respecting the style and contributions of our localization leaders/contributors must be a core part of the approach we take to introducing machine translation into SuMO's KB articles. Accordingly, we're exploring a different approach.
For this issue, the task is to document a new RAG-based LLM approach to machine translation, as well as do some research to add as much detail as possible to the documentation.
We've been exploring the use of Smartling's service as a key part of introducing machine translation into the localization of SuMO's KB articles, but the lack of a straightforward and programmatic way to update Smartling's translation memory with the feedback/adjustments of our localization leaders and contributors is a deal-breaker. Respecting the style and contributions of our localization leaders/contributors must be a core part of the approach we take to introducing machine translation into SuMO's KB articles. Accordingly, we're exploring a different approach.
For this issue, the task is to document a new RAG-based LLM approach to machine translation, as well as do some research to add as much detail as possible to the documentation.