How RWS finds and selects the right experts for Data Services for AI

RWS is widely known as a provider of language technology and services. Not everyone is probably aware that AI is also part of the company's offering, and of course that doesn't just mean machine translation. 

An increasingly important area is Data Services for AI. Among other things, this involves collecting the necessary training data for the machine learning of an AI, preparing it, and optimizing it so that the AI systems can perform the desired tasks. 

In this interview, Nicole Paterlini will give us an insight into the requirements and selection processes to find the experts to help us with this. 

Nicole, perhaps you could briefly introduce yourself? 

Hi, my name is Nicole Paterlini. I am senior manager for Vendor Relations Management at RWS. I am in charge of building our freelance communities that provide translation, localization, and AI services. I started my career as an Account and Project manager but also worked in marketing for a while. This helps me a lot to keep an eye on the different perspectives for a long-term cooperation with our suppliers.

As a language service provider, RWS already employs a large number of experts in all areas of translation. Are the same experts used for Data Services for AI? 

Yes and no. The AI data service includes a wide variety of tasks. For some tasks we rely on the same translation experts, while for other tasks it is sufficient to be of native language fluency or not to have any specific language related knowledge at all. We have for example localization experts who enjoy working on a non-language related machine learning task for a change to break out of their daily routine or language students, who are happy to gain first experiences in the language industry. It allows us to offer working opportunities to a wide range of people. 

What impact does that have on the selection process? What specific requirements are you looking for? 

Due to the large number of different tasks, there are naturally also correspondingly different requirements for the applicants. In general, however, we do a general introduction and a small machine learning test with all applicants. This also gives them a good idea of the type of tasks they will be working on. Depending on the project requirements, additional training or sample tests might be included/provided. Regardless of the candidate’s background, a basic level of English is an important requirement, since most of our communications, instructions and working materials are provided in English.

And once you have applied as an expert, what happens next?

Once we have recruited a new team member, we try to find a project that is as suitable as possible. Since most projects have very specific requirements, the training for each project must of course be individual. Even if our team members are all well qualified, the induction is very important for us. This is how we can achieve the high quality goals in the projects.

What about information security? Do I have to worry about my personal data being passed on? 

No, you do not have to worry. The data you provide us with will only be used for the purpose of training machines. In most of the cases the data is anonymized. However, there are tasks where you might be asked to convey personal information by providing images or audio files of yourself. In such cases we will provide you with the full information so you know how your data will be used and for how long it will be stored. Being GDPR compliant, data protection is a very important aspect to us regardless the projects we work on.

And certainly, also an important aspect is the earnings. How is the work actually paid? Is it worth it? 

I think it is. It all depends on the task requirements and level of difficulty. But in general, our prices are perfectly aligned with the market expectations. And as in all other areas, the opportunities increase as you become more knowledgeable, and you simply create a little more as well. 

Many translators fear that with the increasing use of machine translation there will be less work available for you, or that the work will be paid less. Can tasks in the field of Data Services for AI be an alternative? 

I would not consider AI data service tasks as real job alternative for translators, also because the request for translations in the market is not decreasing, but rather increasing. I see AI as a new exciting area where human input is helping to train machines to do incredible things like recognizing songs or telling me what the weather is going to be tomorrow. It’s an addition to existing business and everyone is welcome to contribute. 

Does this mean that I have to reapply when I am already work as a translator for RWS or who do I have to contact if I also want to work for Data Services for AI? 

If you already signed up with us, you do not need to reapply, you just need to let us know that you are interested to work for Data Services for AI by sending an email to VRM_MachineLearning@rws.com. Our team will be looking forward to getting you up and running in no time. 

Thank you for this interesting insight into the world of Data Services for AI. I am sure that this area will become even more important in the future.