Join top executives in San Francisco on July 11-12, to hear how leaders are integrating and optimizing AI investments for success. Learn More

As enterprises embrace digital transformation, many are expanding customer bases beyond the confines of their pre-pandemic demographics. For example, the cross-border ecommerce market is growing at double the rate of domestic ecommerce — driven by consumers seeking brands unavailable in their home countries. According to Worldpay, 55% of online shoppers worldwide purchased from another country in 2020.

But while digitization provides an opportunity for businesses to expand their target markets, many run up against the challenge of localizing their content for particular customer segments. It’s true that the majority of customers prefer to buy products with information in their native language. But companies, lured by the promise of new business — particularly in the chaos of the pandemic — sometimes cut corners on localizing experiences. Nearly 68% of users in a recent survey said that they encounter web translations that aren’t correct or are confusing because they lack sufficient cultural understanding.

A vast number of companies provide enterprise localization services for enterprises, but an emerging subset is leveraging AI in an attempt to speed up the translation process. San Francisco, California-based Lilt develops AI-powered translation software for marketing, customer support, and ecommerce use cases. Unbabel-owned Lingo24 and Smartling tap a combination of AI-powered translation tools and human translators to localize product descriptions, user guides, websites, software, and apps. There’s also Lokalise, a “continuous localization” platform that helps companies ensure that their software is tailored for target markets.

Language I/O is a small, relative newcomer to the localization space, having raised just $12.1 million in venture capital since its founding in 2011. (That includes a $6.5 million series A led by Omega Venture Partners, which the company announced today.) While Language I/O isn’t in short of rivals in a market that could be worth $5.51 billion by 2028, CEO Heather Shoemaker asserts that the company’s technology significantly differentiates it from other solutions currently available.


Transform 2023

Join us in San Francisco on July 11-12, where top executives will share how they have integrated and optimized AI investments for success and avoided common pitfalls.


Register Now

The challenge of localization

Localization isn’t as straightforward as basic translation. Phrases are longer in some languages than in others (e.g., “buy now” in English is “acheter maintenant” in French), meaning that web designers have to redesign page elements to fit the longer translated phrases. Moreover, some languages run top-to-bottom or right-to-left, which not every website template supports. And cultural considerations in each language can dictate word choice. For example, to a speaker in Japan, where word choice depends on the status of the speaker as well as the listener, friendly and informal ads that play well in the U.S. might rub them the wrong way.

Cheyenne, Wyoming-based Language I/O aims to deliver solutions specifically for customer service localization, enabling companies to use monolingual agents or chatbots to provide support articles, answer emails, and chat across multiple languages. The startup claims that it can get a customer up and running with translations within 24 hours, thanks to the Language I/O platform’s use of AI.

Shoemaker founded Language I/O several years ago, after spending the first decade of her career as a traveling internationalization engineer. After exiting lucratively from a Denver-area startup, she worked with a team to develop the localization technology that now forms the basis for Language I/O’s software-as-a-service platform.

“The need for high-quality, personalized customer support available over digital channels and provided in the customer’s native language has increased dramatically during the pandemic,” Shoemaker told VentureBeat via email. “The expansion of a global remote workforce means that companies are looking to extend the translation and localization capabilities with the current customer relationship management technologies that are already familiar to their support agents. Additionally, [the pandemic] has made travel dangerous and companies are not eager to staff up native-speaking support teams around the globe. This has resulted in a huge demand for technologies such as that provided by Language I/O, which allows companies to get more out of the support agents they already have at home.”

Translation engine

Language I/O employs an engine that intelligently selects AI models for requests and adopts preferred translations for product names, misspellings, acronyms, industry jargon, and slang. Customers flag any mistakes and tell Language I/O which words they want in their dictionary, which enables the models to improve over time across roughly 100 languages.

“[Our self-improving glossary engine] uses datasets generated over the years related to translation quality. Our datasets do not include the actual content sent to us for translation, and this makes us unique in the industry, as our competitors rely heavily on training their models with the content they translate,” Shoemaker said. “Holding onto this content for training is dangerous because personal data is often embedded in support chats and emails — as much as one might try to ‘pseudonymize’ it, you can’t catch everything. Companies get nervous when their customers’ personal data is duplicated and stored outside of their own systems because that’s just one more opportunity for a breach. Instead, Language I/O has gathered translation quality data over the years related to metadata related to each translation request.”

Language I/O integrates with customer relationship management systems including Zendesk, Oracle, and Salesforce and offers an API that allows clients to access company-specific translations. These systems benefit from the aforementioned feedback provided by agents and the professional linguists that Language I/O works with to fine-tune its core technology.

Future expansion

It all depends on the accuracy of the translations — and algorithms aren’t perfect in this regard. But by one estimate, AI-powered conversational solutions can reduce customer service costs by up to 30%. That’s perhaps why 56% of companies say that they’re investing in conversational AI technology to improve cross-channel experiences, according to Deloitte.

While 46-employee Language I/O’s platform is currently focused on translation in channels like email, articles, chat, and social messaging, Shoemaker says the company is poised to extend beyond basic support to “anywhere that businesses need conversational translation.” Think: Slack channels, gamer-to-gamer chats, virtual meeting tech, and learning management platforms.

“The [new] money is to allow us to scale rapidly with the goal of tripling revenue in 2022,” Shoemaker said. “Our product focus this year will be in the field of chatbots and conversational AI and for our machine learning team, which is focused on machine translation quality improvements.”

VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.