The Uber of
Translation platform Fairlingo
Fairlingo’s founder had been active in the world of translation for more than ten years, mainly with his own translation agency. He came to us with an innovative idea for an online translation platform.
What would it need to be? A place where everyone could easily and directly translate texts with a high-quality profile, guaranteed to be done by real human translators. A transparent, modern and, above all, user-friendly translation platform. We were immediately very keen and joined as a strategic and technical partner.
Together we took up the challenge to develop a software product where translator and client come together. A stand-alone platform where the quality and transparency of a translation quote and the process itself stand paramount. An intelligent, versatile and fine addition to the app economy.
- Conceiving and devising a product plan
- Branding, UX and UI design
- Technical platform development using Laravel and React
- AI-driven suggestions system to assist human translators
- Keeping track of platform availability
Every translation assignment is unique, so the Fairlingo platform had to be able to adapt to the demands of multiple industries, assignments and clients from different countries, with many different languages.
A major challenge in developing the translation platform was to design and implement UX that immediately felt intuitive for everyone. Both for clients with a translation assignment, and for translators who complete assignments. We wanted to assist translators using smart tools, so that they could focus fully on the translation assignment.
To ensure that translations are of high-quality, we had to set up a process where other participating translators assess, review and possibly revise translation work. This process had to reward great translations and discourage poor translations.
The platform had to be versatile enough to cope with the many different file formats in which source texts can be supplied. In addition, translations had to be delivered in the same file format, where possible.
Both client and translator want continuous oversight of the translation process from start to finish. That meant always being able to track an assignment in real time in Fairlingo, from quote to final delivery.
Add assignments easily
The user interface in Fairlingo makes it as intuitive as possible for clients to add new assignments. The most important functions are immediately within reach: uploading your document, choosing the source language, the target language and requesting a quote. You will receive the best possible price for the translation assignment within 30 seconds, thanks to the transparent process and smart algorithms that can handle all kinds of file formats.
In addition, you as client can track your translation in real time, including interim versions. You also have the option of selecting the complexity level for your translation. We have added an option for choosing a translator that you have worked with before, indispensable if you want to have certain terminology translated consistently across various translation assignments.
The App Economy
As a translator you can easily sign up for Fairlingo and following a number of verification steps you can start accepting assignments immediately. You indicate your language preferences and immediately have translation assignments you can choose from. This means you also determine your own work pace and number of assignments that you take on.
All translations submitted by a translator are checked by a second translator, to ensure the quality of the translation and to keep the process transparent for the client. Once the translation is approved, payment is sent to your account immediately. It’s that easy.
As you complete more translations, you will not only earn good money with Fairlingo, but you will also earn experience points. Having earned a good number of points, you can gain access to specialist translation assignments which attract a higher rate.
In just a few weeks, we developed a prototype suggestions system for Fairlingo to assist human translators. We did this by training a neural translation model using TensorFlow and the data we had at our disposal.
The system worked rapidly and made good suggestions when translating an assignment. Unlike translation sites, our suggestions system includes some of the text already translated in the suggestion. The autocomplete function automatically adjusts suggestions when translating. A perfect tool for translators. Instead of working in competition, AI now helps human translators. They complement each other, making the translation process even smoother and better.
Incorporating machine learning into a product can bring various benefits, such as a unique selling point compared with other services. For more information on the machine learning model we have developed, and the results from it, we have written a scientific paper (PDF) .
Working closely with Fairlingo's product owner ensures that we are always aware of future development plans.
Being aware of the long-term plans for a product gives us the opportunity to work iteratively on new functionality. Using a flexible development process, we produce beta updates that are tested by the product team at Fairlingo.
Testing new functionality using real translators always benefits the product. This way you get insider knowledge from users that you would not otherwise have been able to incorporate into a new product version.
Do you want
to create a
In Fairlingo, we have created a unique platform that brings together translators and clients with translation assignments. Do you have an interesting idea for an app economy platform, too? Contact Joris for an exploratory chat.