Build customised KantanMT engines using translation memories
Machine Translation is an important tool in your localisation mix, but how do you separate yourself from the poor and high performance engines. Machine Translation can be an efficient way to achieve higher levels of translation productivity as long as your engine is relevant, of high quality, and client-focused. KantanMT helps you easily build high performance machine translation engines for your clients, with little machine translation or technical expertise required.
High Performance MT: Build high-quality engines with our dashboard
KantanTemplates™: Use KantanTemplates to share common training data-sets across multiple KantanMT engines and update them all in one go.
Kantan BuildAnalytics: Use BuildAnalytics to analyse the performance of your KantanMT engines and to spot any data gaps within your training material
The Power of 2! KantanMT builds a TotalRecall™ translation memory to improve both the performance and accuracy of your KantanMT engines
Training Data: Use Translation Memory Interchange files (TMX) as training data
Data Cleansers: Develop your own data cleansers using our Preprocessor technology
Terminology: Use Terminology Interchange Files (TBX) to compliment your training data
KantanLibrary™ Training Sets: Use our training data sets to build larger KantanMT engines
Monolingual Training Data: Use monolingual training data to improve the fluency of your engines
Custom Test Data: Use your own test data to continuously measure quality improvements in your KantanMT engines
Easy Configuration: Manage all your KantanMT engines in one easy to use dashboard interface
KantanMT makes building high-performance machine translation engines as easy as 1-2-3!
Here's what our members are saying:
KantanMT has been used at UCL to teach students at master's level.
R.B. Pinero, Dr., UCL, London
Since signing up with KantanMT, we have been able to take on more work and increase capacity.