Seiler joined the company two years ago as Software Engineer for NLP (natural language processing) after having transitioned from the legal field. (He holds a law degree from the University of Zurich.)
It was while working as a research assistant at university and writing a doctoral thesis on banking regulation that his interest in linguistics was piqued.
Seiler told Slator, “Part of my research was to inform myself on new technologies such as AI that were emerging at the time. This was when I realized that I did not only want to write about these new technologies, but to actually understand and use them myself to solve real-world problems.”
This motivated Seiler to study Computational Linguistics and Computational Science, which then led to his joining TextShuttle as a software engineer in charge of machine translation (MT). In the role, he oversaw the customization of MT systems, “while also covering all legal topics, such as contract negotiations and data protection.”
TextShuttle’s new CEO outlined what they do thusly: (1) provide MT with a high level of data protection in a secure environment of the client’s choice, whether external or on-premise; (2) customize MT to specific corporate wording to sound more natural and familiar to the client; (3) help clients integrate specialized translation systems into a CAT tool of their choice and assist in company-wide rollout.
According to Seiler, a key differentiator is that TextShuttle also incorporates “breakthroughs in MT research into our products much faster than others. To name just two examples, the terminology inclusion mechanism that DeepL offers has been available in our solution since well before they launched it; and our feature to specify the politeness of translations is not only available through the web interface, but also within CAT tools.”
Asked for details on TextShuttle’s shareholders and revenue metrics, Seiler would only say of the former that “they’re all on our website”; and regarding the latter, “Our revenue has exceeded the seven-digit boundary this year. Around half of it is recurring, with an upward trend.”
About the other members of the leadership, Seiler said Samuel Läubli will continue to serve as CTO (“My taking on the CEO position allows him to focus on what he enjoys most: research and development”), and will take the helm from founder Martin Volk as Chairman of the Board of Directors. Volk will remain on the board.
Meanwhile, also on September 1, 2021, Simona Todesco, previously Head of Operations, stepped into the role of COO.
Ideation to Implementation
TextShuttle was founded back in the day when statistical machine translation was “the big thing” and the goal was to produce systems geared toward media, according to Läubli who joined SlatorPod in May 2021.
Today, the company offers (neural) MT consulting — from ideation to implementation — for key clients focused on multilingual communication. This includes banks, insurance providers, language service providers, and the public sector, mainly based in Switzerland but, more recently, Germany as well. According to Seiler, their goal now is to expand business across the entire DACH region.
Tech-wise, Seiler said, they see an opportunity over the midterm for the use of TextShuttle’s core technology in other multilingual communication tasks, such as speech-to-speech translation and automated journalism.
SlatorCon Remote December 2022 | Super Early Bird Now $98
A rich online conference which brings together our research and network of industry leaders.
Always on the Lookout for MT Talent
The TextShuttle CEO described a typical client case study: “A prospective client wants to procure an MT solution adapted to their corporate wording and style guides. The use case of our software is usually two-fold. It is used by the internal language service via CAT tools and as a self service by employees via a web browser.”
Having nailed down specific client requirements, TextShuttle develops a specialized MT system with the client evaluating translation quality several times and giving feedback — which, in turn, is used to improve the MT system. This done, the company helps integrate the specialized MT into the client’s CAT tool, making it available for use by all employees.
TextShuttle’s engineering team comprises a product team (that develops key tech components) and a research team (focused on optimizing MT quality). An operations team, meanwhile, manages MT specialization projects by coordinating with clients. Asked if they are currently hiring, Seiler replied, “We are always looking for talented people in the field of machine translation.”
Given the massive investor interest in AI in general and NLP in particular, has TextShuttle spoken to any investors? Seiler said, “We haven’t been looking for investors — yet. We’re considering fundraising more seriously than in the past, but we won’t be rushed.”