A Look Back at Yandex Tech in 2018
By Mikhail Parakhin, Yandex CTO
2018 has been another exciting year of innovation at Yandex. Ahead of the New Year, I wanted to take the opportunity to recognize the work of our teams and the progress achieved since this time last year.
Over the course of 2018, Yandex increased our engineering staff by over 40% as we set out to create new opportunities for users, businesses, and the global tech community. Together, our growing teams continued to develop innovations that improve the daily lives of our users and help build a better future for them.
This year, our intelligent assistant, Alice, expanded from its initial core functionalities to over 50,000 skills. After first being launched through our search app, Alice is now available on eight different platforms ranging from smart speakers to connected cars, rapidly shifting how Russian users and businesses are interacting with intelligent voice technologies. Alice is revolutionizing the way users interact with their devices across our ecosystem of connected products and services.
In the past year, Alice also developed the power of vision with the integration of best-in-class reverse image search capabilities. This skill helps users better connect with the world around them, whether they are traveling and curious about a landmark or spot a pair of shoes to purchase as a gift.
In December 2017, our self-driving car was tested on closed tracks against winter weather conditions on the outskirts of Moscow. This December, our self-driving cars are actively offering rides to passengers in two tech hubs with just a safety engineer in the front passenger seat. To date, Yandex has provided over 2000 autonomous ride-hailing trips between the two locations and we are now testing the cars in three countries - Russia, the US, and Israel.
While these advancements deserve to be celebrated and will most certainly continue to shift our daily routines in major ways, I am most excited about the advances in AI that are subtly improving user experiences and continuing to raise both our teams’ and our users’ expectations for what Yandex can deliver.
One example of these impressive AI advances in 2018 was the new neural-network based DeepHD super-resolution technology for upscaling photos and videos. In building DeepHD, the Yandex computer vision team created the first video super-resolution technology using GANs that offers users better quality content in real time. As more and more users shift to their laptops or phones for content streaming, in 2018, DeepHD helped users across Yandex’s streaming services rediscover their favorite classics in stunning quality and watch current programs with a significantly clearer picture.
Another example of subtle but important AI innovations this year was the work of the Yandex.Weather team. Throughout 2018, the Yandex.Weather team continued to advance our up-to-the-second weather forecast service by using satellite imagery as a new data source for precipitation maps. Using AI, Yandex improved traditional weather forecasting techniques to provide our users with highly advanced and accurate weather updates. And with the use of satellite technology, Yandex.Weather models can utilize more data and expand their service area.
It’s exciting to think these innovations have quietly changed how modern users expect to receive their weather updates today, when in our lifetimes we amusingly watched Doc tell Marty McFly that the rain would stop in five seconds in Back to the Future. Building and continuing to advance a service that can do just that is, as Doc said, “Right on the tick, amazing. Absolutely amazing.”
With this year’s search update, Andromeda, we continued to find ways to personalize the search experience for our users and help them find answers as quickly as possible. Our search results now provide enhanced quick answers to give users even more information from the results page, and they feature badges that note which sites are the most accurate and relevant, making it easier than ever for our users to find everything they need right from the results page.
Throughout 2018 we also continued to update our open-source gradient boosting library, CatBoost, and our open-source database management system, ClickHouse, which have both seen major adoption by top global companies throughout the year. Yandex also introduced the new intelligent cloud platform, Yandex.Cloud, providing companies with flexible IT solutions using scalable virtual infrastructure, automated services for managing popular database systems, and AI-based Yandex services.
In 2018, we also maintained our efforts to diffuse AI knowledge to the wider tech community, as the Yandex School of Data Analysis introduced several new courses and opportunities for learners everywhere such as a new one-year ML career advancement program in Tel Aviv. Yandex team members also further advanced AI opportunities by publishing significant research for the global community such as a recent paper at NeurIPS, “Non-metric Similarity Graphs for Maximum Inner Product Search.”
While this doesn’t cover everything we achieved this year, I want to thank all the teams here at Yandex for their work throughout 2018, and to the users and businesses for continuing to drive our innovation. We look forward to another innovative year in 2019!
Happy New Year!