The 2017 Speech Industry Luminaries: Rand Hindi
THE PRIVACY PRACTITIONER
CEO and cofounder of Snips
Though voice assistants are still relatively new, voice-activated technology is quickly becoming the dominant user interface for devices of all kinds. That has created all sorts of privacy and security concerns for the industry.
But as Rand Hindi, cofounder and CEO of Snips, has proven with his company’s most recent release, performance and privacy can coexist when it comes to speech technologies.
Hindi has been at the forefront of the development of voice platform artificial intelligence, but for him, user privacy has always been of paramount concern. So it was no surprise a few months ago when his company released the world’s first open-source, end-to-end voice platform offering hot-word detection, speech recognition, and natural language understanding, based on deep learning, that runs entirely on-device, free from the problems associated with the cloud and Wi-Fi. This makes the technology inherently private by design.
Unlike competing products, Snips’ platform does not need to collect massive amounts of data from users. That could be a critical difference as new digital privacy rules requiring more explicit permissions to collect user data go into effect around the world. Additionally, the ability to operate offline can be a game changer for use cases like connected cars, where internet connectivity is not always a guarantee, and in areas such as healthcare or banking, where consumer data protection is mission-critical.
Still, despite operating entirely on-device, Snips claims that the accuracy of its natural language technology rivals similar technology from Amazon, Facebook, Google, Microsoft, IBM, and Apple.
With this latest release, Snips is also democratizing the technology by allowing developers to use its artificial intelligence–driven platform to embed voice assistants into just about any device they make. At the same time, a consumer version will be available on the web. Snips is also selling the voice platform as a component instead of as-a-service, meaning that device manufacturers can purchase the technology and pay just once.
This latest product represents a pivot for Snips.ai, which is based in Paris. Its previous big product release was a consumer-facing iPhone app that used artificial intelligence to create a hyper-smart personal assistant. Here, too, Hindi’s strong approach to privacy played out; the app did all of its computing on the phone rather than sending, storing, and processing it on some far-off cloud server.
And analysts expect a lot more from Hindi, who is just 32 years of age. He started coding at the age of 10 and founded a social network at 14 and a web agency at 15. He showed interest in machine learning at 18 and completed a doctorate in bioinformatics at 21.