How is AI impacting the digital agenda at Uber? And, how is it a critical factor in the digital journey?
We want Uber to be the operating system in everyday lives for millions of Indians and around the world. Achieving this goal requires effortless prediction and optimisation at every turn. At Uber, opportunities range from matching riders to drivers, to allocating jobs to food delivery partners, to suggest optimal routes, to finding sensible pool combinations, and even to create the next generation of intelligent vehicles. All these challenges need to be met at the Uber scale – 700 cities across 63 countries for rides and 500+ cities in 36 countries for eats. Hence, we are combining state-of-the-art artificial intelligence (AI) techniques with the rich expertise of data scientists, engineers, and other users.
AI is at the core of tech at Uber and extremely important in our journey to connect the physical world to the digital world in real-time. Hence, we have an exclusive team within Uber called, Uber AI, responsible for the research and advancements in artificial intelligence to solve challenges across Uber.
How is Uber leveraging AI for predictions and monitoring of customer and driver behaviour, leveraging data to empower the customers?
Uber’s daily operations generate a humongous amount of data, such as the number of trip requests or food orders, which gives us visibility of how and where to improve our services. We apply data science, artificial intelligence, machine learning, and other innovative technologies in our daily processes. For example, predicting the traffic and the patterns over the course of the day to provide our users with better traffic predictions, ETA and hassle-free routes. We also extensively leverage data science and AI methods for personalising experiences, automating processes, scaling customer services, building the next set of innovations such as self-driving, etc.
Machine learning, deep learning, and probabilistic programming have shown great promise in generating accurate forecasts. In addition to standard statistical algorithms, at Uber, we have build forecasting solutions using these three techniques.
How is AI helping you identify and react to problems, and propose the right
services?
Uber is about igniting opportunities around the world through mobility. AI is at the very core of the digital world guiding physical mobility.
For instance, we are driving self-service and resolutions on the app. Through NLP, we would understand customer issues and route them correctly to agents and solve issues efficiently. AI also performs ticket analysis to identify how to best handle these queries and issues, forming an approach of feeding it back potentially into the product to improve user experience and leading to more defect-free products.
The broad streams where we extensively deploy AI to develop state-of-the-art computer vision and machine learning solutions, conversational AI to make all Uber interactions as natural as talking to a friend. Further, leveraging sensors along with algorithms to solve for crash detection, enhanced location accuracy innovations, and even sent phone mounts to drivers for safer driving. We also invest in AI research areas which are very relevant to Uber.
How are you creating the right balance of the skillset to compliment this high-end technology?
We run substantial work in Uber’s machine learning operations for Advanced Technologies Group, Uber AI and core tech groups from India. We also run a Machine Teaching incubator to research on setting up industry-leading practices for ML Ops and building tooling and technology that helps scale machine learning and training data creation across Uber.
The skill sets at Uber tech run deep to broad. We are in a growth phase and hiring the right talent continues. We are hiring at all levels – best campuses and lateral with a high bar. Also, there are many on the job learning opportunities. Besides events, fireside chats, tech talks, workshops, tech fairs employees can opt for a rotation program to work in different teams and locations.