It's a lot for having me. Let me give you some, like three of introduction to myself. So my name's Lay. Graduate from UC Berkeley in 2017 with a PhD in some engineering, and after that I joined the Uber directly on. But I'm country, a data scientist Uber needs focusing on over its marketplace. Prediction. My story is actually pretty simple. My dad wants me to become a medic medical doctor, and he wants me to go to medical school after high school. But at that time, I just wanted to execute my own wheel and explore a new city. So I just pick a random major and know exactly rhythm. But at that time, so it's very pretty random to me on which is transportation engineering. I had no idea what trust British engineer would do, but I thought it should be pretty straightforward because way interacted with transportation every day, right? So my common sense told me, and that's probably a pretty easy major. But it turns out that transportation engineering is a very challenging, very difficult problem and the major, So I become increasingly fascinated with this major that I apply it to him I'm a mess program and the PC program still focusing on introspection, transportation, engineering, Onda. After I graduated, I joined the transportation related company, which is uber, and now I'm still here.Yeah, you know, different. Different doctor.

So I uber under the data science team, we have to sort of jobs related to data. One is a data scientist, and now there is a product analyst or before we call it a data analyst. My job is a data scientist, so we focus a lot building mottoes. Uh, he's any new features, running experiments, ionizing the results, Andi then making model or new, featured all decisions. Um, the typical work hours is about, like, 40 hours per week, but currently we all work from home, so that's sort of there could be more or less depends on your own schedule. But prior to that prior to Kobe, actually, we have very flexible schedules on working from home or in office. Off course they're working from home. Policy varies across teams, but we do encourage everybody to counter the obvious, because person to person chat is often the most effective way for us to communicate ideas. One thing I want to mention is that uber does give employees credits to take over rise and the uber on the or are we overeat. So that's a pretty cool employee benefits. So even though you live pretty far away from the company, you can take your rights

So I said data scientists, We need to deal with a huge amount of data. Right. So you need no Seiko to query data. You also need to know spark in order to be with data pipelines. Onda, uh, uber actually has an internal much learning Pyfrom. So most of our models are trained and deployed on their platform. Eso That's a really effective way for us to scale up their machine learning exercise. But most of our finances and the model part prototyping Adan through python. So you didn't know pie phone on the some. Maybe basic are, um Langridge. Um yeah, that's about it.different ones, so the pattern teaches us more like open source platform where everybody can benefit that, whereas for over internal platform is many used by it by us by uber's.