For the last decade, Deep Learning has been gaining ground in the translation industry. Neural Machine Translation is becoming a powerful resource for both companies that seek to localize their content and LSPs looking to offer new services that allow their clients to power up productivity and budget.

Now, what implications does this have for the Program Managers and Project Managers of LSPs providing Machine Translation Post Editing services? Here are a few tips so all your MTPE projects run smoothly and meet your client’s expectations.

4 Keys to Machine Translation Post Editing (MTPE) Projects

Think Productivity!

From early stages in the quoting process and when planning and executing an MTPE project, bear in mind that one of the perks of offering this service is being able to localize more content, faster. Estimated productivity varies depending on the type of editing your client needs—light MTPE is much faster than full MTPE—but overall productivity is significantly higher for this service than for fully human-translated projects.

Set Clear Quality Expectations

Depending on the type of content your client wants to be localized, budget, and expected TATs, you can opt for Light MTPE—often referred to as good enough or fit for purpose—or Full MTPE—human-like, publishable quality. As a good rule of thumb, always discuss the different possibilities with your client, and decide which option best suits their needs before making any decisions. Remember that not all content has the same visibility, and setting clear expectations in terms of quality, throughput, and costs will drive you forward with fewer bumps on the road.

Manage Terminology from the Start

One of the most common error types found in raw Machine Translation are terminology and nomenclature discrepancies. If you’re a PM planning an MTPE project, make sure you consider glossary creation and MTPE feedback as an essential step in the early stages of the process. This will help you develop a safer Risk Management plan, provide clearer guidelines to your team, and better align with your client’s preferences in the end. At Win & Winnow, we work side by side with our clients to develop, update and maintain reference materials that help create an identity for each brand and set comprehensive quality guidelines for our teams.

Rely on Trained MT Post Editors to Curate your Output

Now that you have carefully thought and planned your MTPE project, it’s time to look for a team of experienced, well-trained linguists who can take care of the different stages and aspects of the process. At Win & Winnow, we rely on our Quality and Linguistic Departments to develop and monitor customized Quality Programs for our clients and assemble our translation teams with a clear set of criteria: field of expertise, training, and years of experience. Working with the right team who can provide feedback on your MT output and curate the final text will ensure the quality of the TM is continuously improved with each project, and make Deep Learning more effective in the long run.

Do you want to know more about our services? Contact us today for a free assessment and quote!