Machine Translation Workflow Integration

PangeaMT is pleased to announce the release of its latest version API to users of its Pangea machine translation as presented in Localization World London 2013. Using an innovative technology developed by our technical team, the REST-based API gives translation teams and translation companies, language service providers and even the savvy freelancer, the opportunity to integrate PangeaMT into translation workflow systems. Thanks to this service, Pangea users can offer customized engines to their clients and either use the service on-demand or monetize on their Translation Memories by creating domain-specific engines with automated cleaning routines and re-training. This technological breakthrough, coupled with our expertise in machine translation deployment and TAUS post-editing guidelines where we participated, make Pangea the fittest solution for large and small MT users who want to “own” their solution. “ Thanks to this new improved service, our clients can achieve faster and better levels of translation automation. This is a massive competitive edge in a market such as ours, already suffering from price pressure and faster turn-around times”. The REST-based API enables PangeaMT users to integrate their customized engines in their workflow, so they can manage the training data of the engines (upload, delete, change and consult the memories used for training) and manage the engines themselves (create, delete, retrain and consult). PangeaMT API can be accessed from any environment that can establish HTTP connections with the PangeaMT server. In API invocations, HTTP headers can be used to set whether feedback data should come back in JSON or XML format. Users can also create API plug-ins for their preferred CAT tool and Pangeanic is working on plug-ins for all major translation environments like MemoQ, XTM, MemSource, SDL, Swordfish, etc. If you need more information about the PangeaMT API or to request an API token please contact us here. What is PangeaMT? Pangea is a cloud-based implementation of Moses statistical Machine Translation technology. Leveraging the power and flexibility of the cloud, PangeaMT effortlessly scales to generate a high-quality, low-cost Machine Translation platform for small-to-medium sized Localization Service Providers. PangeaMT was born in 2007 and has been very successful in its goal of making Machine Translation easy to use, inexpensive and accessible to all. Its famous DIY SMT philosophy set an industry standard in translation automation nowadays demanded by many. We offer the service to large and mid-sized LSPs that are eager to integrate machine translation solutions into their localization workflow. The features are virtually the same as the ones offered by large systems and make small and medium sized LSPs more competitive by offering machine translation to their clients. Many research studies by TAUS, CommonSenseAdvisory, etc point to a large majority of language practitioners thinking about implementing machine translation at some point in their company if they haven’t done so. However, most translation agencies lack technical know-how, hardware or expertise to install, deploy and implement machine translation widely, plus creating language models, training, creating and customizing their engines, running quality tests for themselves or their clients. This is what PangeaMT does for you. We will even run data cleaning routines for your peace of mind. You can have unlimited access and translation, customized engines that learn fast from post-edited material… an intelligent system which will accelerate your workflow and will pay for itself very quickly.

M.Ángeles García

07/01/13