Can AI completely replace human translators? Probably not. AI translations tend to lack the lexical richness of human-translated translations, mainly because AI models make choices based on probability rather than real-world experience. While AI can certainly produce “accurate” translations, those translations will lack the spice of life that a textbook version of the original might have.
For many companies, accuracy is what they want, making the AI translation field an attractive field. Founded in 2016, Smartcat is one vendor that offers automated translation tools for businesses, and Ivan Smolnikov, co-founder and CEO of Smartcat, says business is booming.
“We have over 1,000 customers, including 20% of the Fortune 500,” he told TechCrunch. “Most of Smartcat's customers are large global corporations, but they also include a number of local and international government agencies.”
Before founding Smartcat, Smolnikov was a physicist at the Russian Academy of Sciences, where he studied fiber optic materials. After two years working in the lab, Smolnikov decided to try his hand at entrepreneurship and founded ABBYY LS, a language services company.
In fact, it was at ABBYY that Smolnikov incubated Smartcat, which was spun out as an independent company in 2016. Smolnikov left ABBYY that same year, where he served as a director at the time.
“I founded Smartcat to reinvent the traditional translation agency model, which is based on long, manual, human-driven supply chains that are time-consuming,” said Smolnikov. “Our AI platform supports a wide range of use cases for enterprise customers working with multilingual content.”
Boston-based Smartcat offers tools, apps and managed services that help companies translate written and verbal content, including e-learning courses, websites, files and software, into about 280 languages. Smartcat doesn't necessarily train AI translation models itself, but instead runs content through a “matching engine” to determine the right third-party model for the content and the target output language.
Smartcat says it will fine-tune its translation models, including in-house models trained by customers themselves, when it makes sense, such as when a company has a large database of common translation phrases they want the model to “memorize.”
Acknowledging that AI will sometimes make mistakes, Smartcat also offers access to a network of translators and copyeditors who can help with review for a fee. “Customers can choose from a range of translation options, including AI translation, human translation, or a combination of automatic translation and professional editing,” Smolnikov explains.
A few years ago, some translators on Reddit claimed that Smartcat had a problem with dishonest (and non-paying) clients, but Smolnikov assured me that this issue has been resolved and translators can charge as much as they want minus Smartcat's 2% to 8% commission per payment (the exact commission depends on the volume of work processed and the country where the client and translator live).
While there are many companies selling AI translation services, including human translation services (EasyTranslate, D-ID, DeepL, Lilt, Lengoo, etc.), Smolnikov sees Smartcat competing primarily with legacy translation companies and in-house organizations.
“Traditionally, companies have relied on outsourcing to agencies,” he said. “Insourcing was a different approach, but doing it manually had scalability issues…SmartCat's focus on language AI quality offers practical [alternative].”
Investors seem to agree.
Smartcat announced Tuesday that it had raised $43 million in a Series C funding round led by Left Lane Capital with participation from Lead Edge Capital, bringing the company's total funding to $70 million. Smolnikov said the new funding will be used to expand its 200-person team, support product development and support various ongoing marketing and sales efforts.
“As an early entrant to the market, SmartCat has a broad portfolio of customers and is uniquely positioned to increase the depth and quality of its product offering and maintain a competitive advantage over the long term,” Vinny Puzzi, managing partner at Left Lane Capital, said in a standard statement.
If the most optimistic predictions about the AI translation field come true, these “advantages” could translate into substantial profits (no pun intended): According to Grand View Research, the global market for machine translation solutions could be worth $978.2 million in 2022 and grow at a compound annual growth rate of 13.5% from 2023 to 2030.