Language Business Solutions team and typical Project Management flow Language Business Solutions team and typical Project Management flow
Feb 25


Increasing globalisation contributed to an impressive growth of the translation industry in the last ten years increasing it from $23.50 in 2009 to $46.52 billion dollars in 2018 according to Common Sense Advisory. This growth wouldn’t be possible without help of new technologies, CAT tools, statistical machine translation and now neural machine translation (MT), says the team at Language Business Solutions, gold sponsors at the EUATC's forthcoming annual conference, T-UPDATE taking place in Tallinn 25/26 April. 

Use of AI is already very successful in the terms of MT and allows LSPs to meet the challenges of translating increasing volumes in shorter and shorter timescales. However, it’s not the only field where AI is revolutionising LSPs: project management will be impacted significantly by the introduction of AI in Translation Management Systems (TMS). Rather than machines “stealing” PM’s jobs, AI should see this technology make a qualitative improvement to the tasks that Project Managers perform, making their working life easier and more interesting. Project Managers have often been overloaded with numerous repetitive and administrative tasks. Our simplified illustration shows a typical manual workflow in translation company. At each stage, PMs switch from one software package to another, generating a large number of clicks, writing numerous emails, creating folders, copying and pasting various pieces of information, using spreadsheets work out prices and calculate due dates and so on. Nowadays, thanks to most developed TMS, which recently started to offer automation features, PMs can benefit from the integration of software such as Outlook and CAT tools.

Automation can simplify administrative tasks and perform it faster than PMs would. For example: automatic calculation of exact time and budget of the service (quick preparation of quote), automatic folders creation (to avoid any inconsistency in folder’s names created by different members of the team), preparing detailed statistics (to track easily a company’s objectives) or even automatically assign the job to suppliers in case of repetitive projects (with automatically generated email).

Automation is very useful since the algorithm helps to optimise the steps where a PM isn’t necessarily being asked to add value. Moreover at the same time it decreases their administrative workloads by around 30%. TMSs are evolving constantly by making their interfaces more user-friendly, making functionality more intuitive, automating not only production activity of the company, but also CRM or accounting functionalities. Still, the automated functionalities have to obey a set of rules before making it efficient.

AI in TMS will be going beyond data integration and automation of repetitive work. At the moment TMSs are fed by humans with pre-programmed rules, but future TMS will be self-driven softwares able to understand projects and content, and maybe able to enter in contact with members of the team. 

One of the possible visions of future TMS skills could be speech recognition. With this a Project Manager could simply ask questions to the chatbot: “What is the availability of translator X?” Another evolution of TMS could be recognition of the content of source files allowing more accurate selection of the linguist. It may also be an automatic project creation adapted to a client’s needs and instructions. AI could, potentially, provide Project Managers with solutions based on data-learning from previous projects.

Let’s imagine a new client coming with a large project (with tight deadlines), which should be divided between many translators in rolling delivery. AI would be able to propose the most efficient solution built on previous experiences (divide or not divide, if yes in which proportions and between which suppliers, which tools?). This predictive skill of AI, built on deep learning, leads to automated behaviour that will be efficient. This innovation will change the organisation of translation companies: TMS won’t be only the PM’s assistant, it will participate actively to the workflow. Thanks to machine learning AI will probably take decisions with minimal or light touch involvement of Project Manager.

Today in the translation industry, Project Managers are asking themselves what their future is likely to be. It is true that AI allows you automate processes and reduce manual work, but human skills such as judgment, common sense, coaching, interpersonal, strategic and proactive abilities will not only remain very useful but also essential. 

The nature of PM’s job will change. They will be able to focus more of their energies on different tasks such as creating and maintaining the relationships with clients and suppliers. Furthermore, PMs could also participate more in linguistic tasks, since most of them are linguists. Gradually, AI will generate workflow and the PM will become what post editors are for machine translations: an experienced supervisor of machine’s performance. 

The future of TMS is certainly through AI. The question is when will it occur, and how fast will PMs see their lives change.

Meet the Language Business team at T-UPDATE in Tallinn 25/26 April.