Did you know that English is spoken only by a quarter of the world’s population? This leaves an untapped market that businesses should consider more seriously. However, without a digitally advanced, and scalable language translation strategy, businesses can find themselves limited to specific geographic regions.
Could you tell us what personal experience inspired you to create a service intelligence where humans and machines work together to ease language barriers?
I’ve always been fascinated by AI and language. My mother was a professor of linguistics so language has always been deeply rooted in our household. I started coding at the age of 6 and everything came together very early on. I became obsessed with the idea of building things.
At age of 15, I decided to start my own company. I did my masters and PhD in natural language processing, and during my doctorate program, I took on an internship at Google (which was great, a bit like La La Land for developers). Afterwards, I started thinking about how I could build something new and impactful. I dived deep into exploring how humans and machines can communicate to ensure language translation confidence and trust in all company-customer interactions. My founding of Unbabel was a natural extension of that.
As a language operations platform provider, what do you think are the biggest challenges that multilingual customer service operations face with?
The pandemic put unprecedented pressure on organizations of all shapes and sizes to do more with less. Many experienced an increase in demand for products or services but could not staff up quickly enough to meet that volume. Unfortunately, most organizations today still rely on a siloed approach to language. For example, the marketing team might use translation services, while the sales team hires native speakers, and the customer service department staffs up with native-speaking business process outsourcing (BPO) resources.
This approach is limiting − there is no consistent language strategy that spans an entire organization, which consequently leads to massive inefficiencies. Language operations leverage machine translation with human editors to help create efficiencies for existing customer service agents and help teams focus on hiring for expertise, rather than language.
At Unbabel, you use AI-powered translation solutions to help companies ease their interaction with a customer in any language. How does this technology help to alleviate challenges with language barriers?
We are introducing a new category to the market: language operations. This is a cross-disciplinary function that helps global businesses communicate more effectively with their multilingual customers and stakeholders. It has also accelerated the trend towards digital transformation as well as a shift toward a “rise of operations-for-everything” era.
Language operations (LangOps) is AI-powered, but the initial translation work is reviewed and approved by humans. This blended approach of providing AI-powered, human-refined translation enables more efficient, effective, and authentic translations. It eliminates the need for native speakers in every market but still provides high-quality translations.
Can you share an example of how this may play out in real-time?
Last year clearly created a surge in demand for remote work solutions. Customer service tickets dramatically increased for many companies around the world. Unbabel’s supported Logitech, an international manufacturer of computer peripherals and software, saw an increase in customer service tickets by up to 300 percent. With Unbabel, the company was able to reduce ticket response times from two days to twelve hours effortlessly — supporting language translation for 13 verticals across 16 languages. This helped them raise their Net Porter Score (NPS) by 58 points.
What does this mean for companies on a strategic level? How should they employ and train their contact service representatives?
The most strategic and ultimately the most successful operations teams (both for language and other essential functions) will use technology to eliminate silos and drive collaboration across one or multiple departments.
We are at a tipping point where AI/ML can drive economies of scale for language. Contact centre representatives can connect with customers speaking a more diverse set of languages with the help of AI. This consolidates the function of language under a single team for the entire organization — allowing the organization to scale into new markets quickly.
Let us end this interview with your perspective on the human side of technology. What is your vision for Unbabel’s contribution to the world in times of crisis?
We are committed to removing language barriers by enabling trustworthy and seamless communications between people who understand and need to be understood. This could not be more critical than right now as the pandemic continues to impact our daily lives. Our goal is to give a local experience for an increasingly globalized world.
Given that we are passionate about global cultures and languages, the hate crimes against minority communities fuels our mission to build a universal understanding and unity against hate of any kind to anyone. We will continue to explore new opportunities to innovate and collaborate in the emerging LangOps space to support and empower individuals as well as organizations to digitally connect around the globe regardless of the physical distance between us all.