How Qwant Aims to Bring Privacy to the Search Engine Game with AI

On a PC, a phone or from TV, people want information. They want it at home, at work and on the move. And they want it 24/7. They also want, but seldom get, their privacy along with it.

The upstart search engine Qwant is using AI, powered by NVIDIA technology, to change that game.

The Franco-German startup, which launched in 2013, doesn’t collect users’ personal data. And it doesn’t use cookies to identify people through their browsing activity. This means users won’t have their private browsing information sold to advertising companies.

Qwant search engine on mobile
AI is at the heart of Qwant’s business model.

For most search engine users, remaining anonymous online is near impossible. Search activity and personal data are collected and compiled to tell a comprehensive history of everything people have ever searched for. This can create a detailed timeline that stretches back years.

Advertisers can use this data to target the ads people see, with sometimes disconcerting effect. Shopping for shoes online can result in footwear ads stomping across every page you visit weeks later. And search engines can use the data to adjust search results, potentially changing how a person views the world.

Qwant takes a different approach, making its money using a pay-per-click model. This means it needs highly accurate, relevant search results. And this is why AI, accelerated by the NVIDIA DGX-1 AI supercomputer, has found itself at the heart of Qwant’s search engine.

Qwant uses the DGX-1 to supercharge its deep learning applications, which require very fast analysis and processing of enormous amounts of data, and to make sure its users get highly relevant results. With deep learning, Qwant indexes and ranks what it finds on internet pages, and can better automate its understanding and classification of the content.

This lets Qwant, for example, more efficiently detect and remove spam from search results. And it can detect mature or inappropriate content, whether text, images or video, and remove it from results when the “Safe Search” feature is selected. With a single unit offering the equivalent of 250 conventional servers, the DGX-1 can speed these automated filters across vast datasets.

“Because it combines both powerful hardware and optimized software that we can build upon, the DGX-1 allows us to truly unlock our R&D efforts,” says Eric Leandri, president of Qwant.

Take the Qwant search engine for a spin at https://www.qwant.com/, and leave your ID in your wallet.

The post How Qwant Aims to Bring Privacy to the Search Engine Game with AI appeared first on The Official NVIDIA Blog.