CONSULTANCY - Linguistic technology support Consultant -SSA-2026-HRCS-01
Organizational Context
LSP completed a machine translation (MT) retraining in 2024-2025 for its customized models in five languages. These models require regular retraining based on new texts translated since the last time. The download of bitexts in the five languages required for the training, as well as their cleaning are extremely time-consuming tasks and are done mainly manually. LSP seeks to automatize these processes through the development of a specific tool that would download bitexts from LSP CATT tool (eLUNa) and clean them for training purposes. The models will need to be retrained as well. Additionally, LSP would like to further explore the use of Microsoft applications and technology more broadly to develop ad hoc language support tools to automate specific tasks related to translation, editing and publishing processes. These tools should be aligned with WMO IT infrastructure and require no extra purchase cost. This would reduce considerably manual effort. The selected consultant could also provide advisory services related to the future job management system and any on-going technology projects.
Job Purpose
To support the Language Services Section (LSP) in retraining machine translation models and developing language support tools, leveraging expertise in language technology and Microsoft applications to automate processes and improve translation workflows.
Responsibilities
['Develop a tool to automatically download WMO bitexts from eLUNa for specified language pairs (English > Arabic, Chinese, French, Russian, Spanish) and years (2024-2026).', 'Create a database of downloaded bitexts for MT training.', 'Develop a tool to clean up downloaded bitexts by removing duplicates, same language matches, empty targets, and misalignments.', 'Retrain Microsoft Translator models for the specified language pairs and their reversed counterparts.', 'Evaluate retraining results using automatic scoring (BLEU/COMET) and provide reports.', 'Conduct human evaluation of retrained models in cooperation with WMO linguists and provide reports.', 'Prepare bilingual glossaries based on WMO terminology and acronyms for integration into Microsoft Translator models.', 'Retrain Microsoft Translator models using the prepared glossaries.', 'Evaluate retraining results after glossary integration and provide outcome reports.', 'Take corrective actions or research alternative solutions if retraining results are unsatisfactory.', 'Train the Linguistic Services Support Officer and relevant staff on model retraining and provide guidance materials.', 'Present retraining results to LSP stakeholders.', 'Document the procedure for model retraining and evaluation.', 'Assist LSP by developing language tools for translation, editing, and publishing processes (e.g., cleanup of Word files, validation of references and hyperlinks, glossary alignment).', "Amend WMO template macro ('WMO Ribbon') for integration with SharePoint lists and session websites.", 'Provide support in testing functionalities of ongoing technology projects and advise on their acceptability.', 'Train the Linguistic Services Support Officer and other stakeholders on the use and maintenance of developed language tools.', 'Prepare detailed documentation and guidance material for relevant processes.', 'Carry out other relevant duties as required.']
Work Experience
Proven experience in training and retraining machine translation models (Microsoft and other engines), developing language tools using macros, scripts, and IT scripting tools (e.g., Python), and developing and managing databases/record sets.
Skills
['Knowledge of training and retraining machine translation models (Microsoft Translator required, other engines a distinct advantage).', 'Advanced knowledge of Microsoft applications (Azure, SharePoint, Forms, Power Automate, VBA, etc.).', 'Expertise in Python and similar programming languages.', 'Ability to transfer knowledge through documentation and training.']