December 21, 2017 | Written by: Danny May
Categorized: AI | Client Stories | FinTech | News & Events
Share this post:
Try to picture this scenario: you are in a foreign country, let’s say China, on a business trip. You don’t speak the language, don’t have any contacts, and are relying on a vanilla translation app on your phone to help you get out and about. Then, something unexpected happens: your passport gets stolen. You rush to the police station, try to communicate with the officers… and let’s just say it doesn’t go very well.
Believe me, I understand how you would feel – that person losing their passport in China was me, just a few years ago. That whole debacle, however, was precisely what inspired me to found Lingmo.
Seeking international opportunities
These days, whether you are a small company, a large firm, or a global enterprise, having an international outlook is crucial. Every time your business starts operating in a new country, you open up new revenue channels and unlock new opportunities to sell to a larger, more diverse market.
The reason more companies don’t do this often just comes down to practicalities. Perhaps you don’t know anyone who lives in the country that you want to expand into, and perhaps none of your employees can speak the local language. If you send someone out there to build contacts and sign contracts, it’s important that they don’t feel like a fish out of water. The more at home they feel, the more likely they are to make good decisions and confident deals.
Understanding the context
Translation software could be a key tool to help businesses operate internationally, but in my experience, traditional translation apps don’t actually help much in real world situations. They miss too much of the context and nuance, and they’re generally hopeless at dealing with dialects – which is especially problematic in countries like China, where the language has many different regional variations.
I figured that machine learning and artificial intelligence could be instrumental in building a more advanced, nuanced, and precise translation tool. These technologies, however, need to be trained and tested on large data sets, so we needed to keep feeding our newborn translation software with huge quantities of voice recordings.
We needed to find a nimbler approach that would help our machine learning models get smarter, faster. That’s where IBM Watson came in. By harnessing Watson’s AI capabilities, we realized we could achieve our goals much more efficiently.
By enabling us to train our models on text instead of voice samples, Watson has slashed the amount of data we need to feed in, and accelerated our training cycles by up to 50 percent. This helps us to refine our word recognition and translation capabilities much more quickly in response to user feedback.
Read more about how Lingmo used IBM Watson to move to a cognitive approach, leveraging Watson to enhance speech recognition and train models much more quickly, even with relatively small volumes of data.
Fresh ways to communicate
By injecting the Watson cognitive capabilities into our Translate One2One earpiece, we can now translate between nine different languages with at least 85 percent accuracy in real world tests. Translated content is available to users in near real time, directly via the earpiece and without the need to use a mobile phone.
Thanks to Watson, our app genuinely adapts to real world situations: for example, we have a Spanish professional soccer player who recently came to play for a team here in Sydney. When he arrived with his family, his wife couldn’t speak any English, so we loaned them a prototype of our next product. Now she feels much more comfortable about leaving the house, taking their baby out, and generally making a life for her family in a new community.
We also believe that our app will pique the interest of the travel and transportation sector – particularly airlines, who need to help passengers from all over the world reach their destinations without trouble or stress. In the past, airlines tended to hire multi-lingual staff, but as margins get tighter this is becoming increasingly challenging. With our software, we offer airline staff the ability to understand nine languages instantly, which becomes incredibly useful when they need to assist a passenger in need.
By working with IBM, we have succeeded in developing a cognitive translation tool that truly empowers people, and enables them to communicate more easily – a key milestone in our quest to make language barriers a thing of the past.
Learn more about Lingmo
Read the Lingmo IBM Watson Case Study here
Learn more about IBM Watson Services on IBM Cloud