4 October 2011 went down in tech history as the day Apple presented "Siri". In one fell swoop, artificial intelligence was suddenly available on new iPhones, and as such available to the general public. Other IT companies from California soon followed suit. Microsoft Cortana, Amazon Alexa and Google Assistant have been finding their way into a growing number of households. Yet voice assistants need not be tied so closely to their makers' interests – and in particular should provide enhanced privacy protection. That was the idea by the principals at SemVox when they founded the startup in Saarbrücken, Germany, back in 2008.
Their software is called ODP S3. SemVox hasn't developed a single assistant, but rather an entire assistant framework. Other companies and organizations can integrate this framework into their products, creating their own voice recognition systems.
"A surgeon in a hospital can control the medical technology without having to put down his scalpel and swab," explains SemVox Communications Director Michael Bruss. Or car makers can give their own distinctive personality to their in-car systems and no longer rely on "Standard Siri." Because ODP S3 is customizable: the names, voices, speaking habits as well as all commands and reactions can be individually defined.
As with all intelligent voice assistants, SemVox software technology uses self-learning artificial intelligence. "If I commute weekday mornings and evenings between the same locations, then the system understands where I work and where I live," Bruss says. Sit in your car and issue the command: "Take me to the office," and the navigation system calculates the route immediately. This isn't anything groundbreaking in and of itself, but ODP S3 can understand more than just voice commands: "If the user at some point in the future is driving a car with very good gesture recognition, she can point at a building and ask 'What's that?' and the voice assistant will tell her," Bruss suggestions. The framework has the infrastructure to take advantage of the statement, gesture and GPS data to call up the necessary information — and is built to let companies implement that kind of feature.
In pursuing flexibility and next-level gesture recognition, the Saarland-based company is following an approach similar to that of the industry's heavyweights. Viv Labs, an American startup, is also working on an assistant that will be implemented with open software with gesture controls. The company is backed by the original founders of Siri.
One of the important benefits of ODP S3, however, doesn’t have anything to do with its functionality, per se. "It's all about security and privacy protection," Bruss says. Well-known voice assistants like Alexa always have to connect to their maker's Cloud before someone speaks to them. It's the only way for them to process information. That means that there is no control over what happens to the voice recordings if they're sent to a server somewhere abroad. It's a risk for users that has repeatedly drawn the ire of the German Consumer Protection Board : "Everything you say may be wandering into US Cloud servers right now."
Users of ODP S3 can decide for themselves how and where information is processed: "If a hospital wants to put the server in its own cellar, then that's fine and no information leaves the building," Bruss explains. "But with that said, we can also work completely in the Cloud or on a hybrid model."
SemVox is based in Saarbrücken and currently counts just 50 employees, but stands eye-to-eye with the industry's luminaries when it comes to technology.
The startup is heading to CeBIT in search of partners and customers. This will be the company's third appearance in CeBIT’s dedicated pavilion for startups, which is called SCALE11 and gives "young innovation drivers," (as Bruss puts it) an ideal platform. Investors can leverage this opportunity to get a closer look at the company. After all, only the surface has been scratched on up-and-coming B2B market for voice assistants.
Please click here for more information on SCALE11