About this role
You will be part of a team known for its industry–‐leading machine translation and translation automation systems for localizing software products, documentation, websites, and marketing assets. As such you will be responsible for ongoing development of our machine translation environment and other natural language processing technologies within the platform.
Responsibilities
Develop and integrate Moses–‐based MT engines in Autodesk’s translation platform.
Run regular analyses of raw and post–‐edited machine translation output. Improve analysis automation.
Use translator feedback and MT usage statistics to continuously improve MT quality.
Take part in the design and execution of MT productivity tests. Investigate options for running productivity tests under production conditions.
Participate in natural language technology research within Autodesk.
Implement new natural language technologies and solutions as required.
Evaluate, configure and maintain MT engines from external providers and develop process automations.
Company:
Autodesk
Qualifications:
Requirements
M.Sc. or M.A. in Computational Linguistics or Computer Science with a specialization in Natural anguage
Processing
Familiarity with the Moses SMT toolkit
Experience with object oriented programming in languages such as Java or C++
Experience manipulating large corpora
Ability to stay current with relevant technology trends and developments
Ability to collaborate and communicate effectively with local and remote team
Ability to quickly adapt to new situations and to learn new technologies
Strong analytical and problem–‐solving skills
Preferred qualifications and skills;
Knowledge of localization practices and tools
Experience with functional and reactive programming
Experience programming with Scala, Akka and Spray
Experience working with NoSQL and relational databases
Experience with research in the field of Computational Linguistics
Experience with Agile methodologies
Educational level:
Master Degree
How to apply:
Please mention NLP People as a source when applying