Boosting Player Support Productivity With AI Machine Translation
From Siri and Alexa to the automated chatbots that help us shop or do our banking online, AI has become part of our day-to-day lives. Why then are we still so fearful when it comes to introducing it into professional spaces to help break down communication or productivity barriers?
It’s true that the holy grail of instant and fully coherent AI machine translation may still be further down the road, however, professional machine translation has moved past the deficiencies of consumer-grade, free-to-use AI translation software and programs and observed leaps and bounds in bringing AI assistance to bear on a range of everyday productivity scenarios – including the player support process!
The key to transcending last-gen translation systems is in how AI machine translation functions. Where previous machine translation applications learned by referencing a translated dictionary and applied literal translations to every word, new computer ‘minds’ learn just like a neural network in the human brain. That’s where neural machine translation comes in.
Neural machine translation (NMT) allows information to be pulled from a webbed network of possible translations, comparing the initial source text to previously translated text, building a custom translation tailored within a specific or pre-determined context. This progress in neural machine translation has allowed businesses worldwide to ensure high-quality translations are made at a fraction of the cost and time of manual translation projects.
When looking at this through the lens of a potential player or customer support project, the benefits for team and process augmentation can be understood loud and clear.
Strength in Speed
The first advantage in leveraging neural machine translation for player support teams is that the grunt-work of the translation is, as you would expect, done by the machine. By working in tandem with humans, AI translation software can provide reasonably accurately translated tickets, articles, and responses to boost even the best team’s workflow and turnaround times.
Since the heavy lifting in creating multilingual replies, macros (‘stock’ ticket responses and resolutions that can be sent en masse), and knowledge bases is in the production time it takes to translate a given text into each target language, support specialists often need to take valuable time away from solving customer issues to handle rote and repetitive translation tasks.
With the time-consuming task of translation already handled by the AI, you’ll be freeing up your multilingual support specialists to do what they do best, polishing community posts, creating content, conducting quality control tests, and taking care of your players. (Though it should be noted, AI NMT does still struggle in translating creative content, a fact teams should be made aware of and keep vigilant around.)
This applies equally in recruitment cycles when it comes to upscaling your current efforts or introducing new language combinations/teams. Ensuring your team has access to training information, operations guidelines, and other workflow documentation in their native language also becomes a whole lot faster, meaning that you can set up infrastructure for new team members in a fraction of the time it would traditionally take – a huge advantage when upscaling a project at speed or diversifying your language spread on a deadline.
Cost & Scalability: Opening the Global Gates
With the AI performing near instantaneous translations and your team simply performing QA and QC checks, you’ve already cut down on recruitment and staffing costs immeasurably. And with fewer support specialists overseeing the translation process, you also cut down on the number of people needed on shift at any given time. Good news for any shift manager.
Additionally, a significantly lower translation workload on your player support specialists means each individual specialist can become more flexible and dynamic in terms of the workload or skillset they’re able to take on. This is where neural machine translation can really work its dynamic magic.
There are less obvious benefits to scale presented here also. It is sometimes the case that to install a native-speaking team, companies often need to open new office locations in-country for a specific target language – tying them unnecessarily to a single time zone or set of regional tax or corporate legislation. With machine assistance, companies can break away from geographical and time-zone dependence (and the trials of sourcing expatriate native talent), allowing them to locate their teams as they please.
Quality & The Human Touch
While great advancements in NMT and linguistic AI programs are being made constantly, it should go without saying that the quality produced is still no comparison for true native-level human translation. Processes still need that human touch to ensure that your communications uphold the level of quality you need.
For example, not every ticket a team receives will be written in a way an AI will be able to understand. A player could be using text language, colloquial slang, or simply have low level of grammar or spelling skills. With time, AI can be taught to compensate for and even understand these roadblocks, but they will be no match for a human mind that is looped into how a living language works or can use some quick critical thinking to decipher anything that may be misspelled or confusing in the text.
Efficient AI translations do mean a smaller and more focused translation team is required, but they also call for stricter monitoring in quality assurance processes to ensure consistency, accuracy, and the (albeit decreasing) margin for error associated with AI. Competing with an experienced all-human translation team is no easy task, the AI is always going to struggle in communicating the kinds of understanding, nuance, and empathy that a world-class player support team will bring to the table.
That being said, neural machine translation is an excellent tool for improving your processes, keeping your bottom line tight, and making you better at what you already do. While we are still some time out from AI catching up with humans to the point we can rely on it 100%, NMT and AI translation can be an extremely useful tool in scaling costs, team size, and productivity in a pinch.