Uber built Michelangelo, our machine learning platform, in 2015. Three years later, we reflect our journey to scaling ML at Uber and lessons learned along the way.
Uber developed Michelangelo PyML to run identical copies of machine learning models locally in both real time experiments and large-scale offline prediction jobs.
To improve our maps, Uber Engineering analyzes customer support tickets with natural language processing and deep learning to identify and correct inaccurate map data.
One-click chat, the Uber driver app's smart reply system, leverages machine learning to make in-app messaging between driver-partners and riders more seamless.
Uber's Advanced Technologies Group introduces Petastorm, an open source data access library enabling training and evaluation of deep learning models directly from multi-terabyte datasets in Apache Parquet format.
Uber Eats engineers describe how they surface restaurant recommendations in the app using multi-objective optimization to give eaters the most satisfying experience while maintaining the health of the Uber Eats marketplace.
In this article, we provide a general overview of how our teams leverage forecasting to build better products and maintain the health of the Uber marketplace.
Uber built the next generation of COTA by leveraging deep learning models, thereby scaling the system to provide more accurate customer support ticket predictions.
As powerful and widespread as convolutional neural networks are in deep learning, AI Labs’ latest research reveals both an underappreciated failing and a simple fix.
Uber developed its own financial planning software, relying on data science and machine learning, to deliver on-demand forecasting and optimize strategic and operations decisions.
Curious about what it is like to traverse the high-dimensional loss landscapes of modern neural networks? Check out Uber AI Labs’ latest research on measuring intrinsic dimension to find out.
Applying hardware acceleration to deep neuroevolution in what is now an open source project, Uber AI Labs was able to train a neural network to play Atari in just a few hours on a single personal computer, making this type of research accessible to a far greater number of people.
Differentiable Plasticity is a new machine learning method for training neural networks to change their connection weights adaptively even after training is completed, allowing a form of learning inspired by the lifelong plasticity of biological brains.
Uber AI Labs introduces Visual Inspector for Neuroevolution (VINE), an open source interactive data visualization tool to help neuroevolution researchers better understand this family of algorithms.
Interested in accelerating your career by tackling some of Uber’s most challenging AI problems? Apply for the Uber AI Residency, a research fellowship dedicated to fostering the next generation of AI talent.
Uber Engineering created Omphalos, our new backtesting framework, to enable efficient and reliable comparison of forecasting models across languages.
Uber ATG Toronto developed Sparse Blocks Network (SBNet), an open source algorithm for TensorFlow, to speed up inference of our 3D vehicle detection systems while lowering computational costs.
In this article, Uber Engineering introduces our Customer Obsession Ticket Assistant (COTA), a new tool that puts machine learning and natural language processing models in the service of customer care to help agents deliver improved support experiences.