Uber ATG Machine Learning Platform组招聘多名工程师
122792
Uber ATG（无人车）有多个工程师职位
招人的不是我 ：）
是Uber ATG一位hiring manager。
Looking for Experienced Engineers on Machine Learning Platform Team (multiple openings).
1. 能否介绍一下 Autonomous Vehicles (AV) 大体做什么？
Cars have multiple sensors that gather data (what the car sees in the world), such as camera, lidar etc. Human then label "interesting" snippets. The data is then used to train models offline that perform tasks such as perceptions, prediction, localization, motion planning and control. A single binary with models built into various parts of the system, run on very powerful computers in the car, to drive it around. But before that, hierarchies of tests are performed thoroughly from single ML models to cascades of models, as well as parts/whole of systems both in data centers (pure software tests) and with hardware-in-the loop. The car will perform track tests in special facilities, before they go to public roads. We also test the system with both miles driven on the road and in data centers (simulations).
The technical challenges in the AV world is complex, diverse and fascinating.
2. 新招的职位，在Machine Learning Platform组具体做什么？
Offboard to onboard test: large scale offline testing of software systems and machine learned models, both with pre-recorded sensor logs from the car, and simulation.
Optimizing models for onboard inference speed.
You will work with PBs of data. Your compute happens in a hybrid cloud system (both on-prem and cloud).
3. 新招的职位，不做什么？
While we will get good exposure to all the ML in the Autonomy Vehicle stack, we will NOT be building models from scratch, create new model architectures, etc. This is a platform / engineering team. Our mission is to enable scalable/faster/cheaper/safer/reproducible systems to support research and model/system development.
4. 你认为Machine Learning Platform组的工作，什么地方最吸引人？
You will get to see most of the AV stack. We work with almost every single team at ATG.
The team is still growing and you will have opportunity to build new systems from scratch and own a significant scope.
Depending on your expertise/interest, you can work with people from diverse backgrounds: autonomy engineers, researchers, infrastructure engineers, embedded systems and distributed system engineers.
You will have significant autonomy in the daily work and we are super open to your creative solutions to the challenging technical problems we are solving here.
5. 这些职位需要什么样的背景和经历？
6. 这个职位在什么地方？
San Francisco。
7. 近期Uber裁员。如果我加入ATG，是否暂时不能办绿卡了？
其实没有影响：
所以对现在考虑Uber的人来说，绿卡申请无特殊影响。
8. ATG为什么一边裁员一边招人？
9. 是否招New Grad or intern?
Machine Learning Platform组目前只能直接社招 experienced hire。
无论是New Grad 还是Intern，Uber有统一流程，你必须从Uber网站 -> University Recruiting里提交。
提交简历给我们也没用。
10. 能否分享一下你对于ATG的个人感受？
ATG (as any company) is not perfect. There are definitely many areas for improvement. However a few things stand out as uniquely "cool" for me.
There's only lightweight processes and relatively little bureaucracy. Engineers spend most of our time doing the work. Decisions are often made quickly to move things along.
ATG is very much engineer driven - not PM or data driven. (there could be pros and cons here). Solid technical work counts A LOT.
We do work hard but there are no crazy on-call schedules - there's virtually nothing that requires us to wake up at nights and weekends.
This is a highly technical place full of geeky people. People will talk about tech with you, WAY more than gossip.
11. 如何申请？
本帖隐藏内容需要登录后才能查看。
Preview
招人的不是我 ：）
是Uber ATG一位hiring manager。
Looking for Experienced Engineers on Machine Learning Platform Team (multiple openings).
1. 能否介绍一下 Autonomous Vehicles (AV) 大体做什么？
Cars have multiple sensors that gather data (what the car sees in the world), such as camera, lidar etc. Human then label "interesting" snippets. The data is then used to train models offline that perform tasks such as perceptions, prediction, localization, motion planning and control. A single binary with models built into various parts of the system, run on very powerful computers in the car, to drive it around. But before that, hierarchies of tests are performed thoroughly from single ML models to cascades of models, as well as parts/whole of systems both in data centers (pure software tests) and with hardware-in-the loop. The car will perform track tests in special facilities, before they go to public roads. We also test the system with both miles driven on the road and in data centers (simulations).
The technical challenges in the AV world is complex, diverse and fascinating.
2. 新招的职位，在Machine Learning Platform组具体做什么？
Offboard to onboard test: large scale offline testing of software systems and machine learned models, both with pre-recorded sensor logs from the car, and simulation.
Optimizing models for onboard inference speed.
You will work with PBs of data. Your compute happens in a hybrid cloud system (both on-prem and cloud).
3. 新招的职位，不做什么？
While we will get good exposure to all the ML in the Autonomy Vehicle stack, we will NOT be building models from scratch, create new model architectures, etc. This is a platform / engineering team. Our mission is to enable scalable/faster/cheaper/safer/reproducible systems to support research and model/system development.
4. 你认为Machine Learning Platform组的工作，什么地方最吸引人？
You will get to see most of the AV stack. We work with almost every single team at ATG.
The team is still growing and you will have opportunity to build new systems from scratch and own a significant scope.
Depending on your expertise/interest, you can work with people from diverse backgrounds: autonomy engineers, researchers, infrastructure engineers, embedded systems and distributed system engineers.
You will have significant autonomy in the daily work and we are super open to your creative solutions to the challenging technical problems we are solving here.
5. 这些职位需要什么样的背景和经历？
- 需有良好engineering 基础，工业界项目经验. You have to be a solid engineer first.
- Tech Industry full time working experience required: 1+ year if you have a PhD; 2+ year if you have MS/BS.
- Must have BS degree in CS／CE／EE.
- Bonus: Experience with building distributed machine learning systems
- Bonus: Experience with embedded systems, model optimization
- 扎实的DL／ML 基础: such as computer vision, sensor fusion,
- Graduate degree strongly preferred.
6. 这个职位在什么地方？
San Francisco。
7. 近期Uber裁员。如果我加入ATG，是否暂时不能办绿卡了？
其实没有影响：
- CEO 说这是近期第三波也是最后一波裁员，参考techcrunch.com。
- 裁员前后，绿卡政策相同：入职6个月后可以开始办绿卡
- 裁员的直接影响是 04／14／2020 之前 不能file perm，考虑到面试和H1B Transfer都需要时间，还可能被RFE，等你启动绿卡申请，肯定是4月14以后了。
所以对现在考虑Uber的人来说，绿卡申请无特殊影响。
8. ATG为什么一边裁员一边招人？
- 裁员比例非常小，ATG裁员的情况全部是performance based
- ATG仍然在增长，仍然有很多有趣和有深度的技术问题等待我们探索
- 裁员已经结束了。
9. 是否招New Grad or intern?
Machine Learning Platform组目前只能直接社招 experienced hire。
无论是New Grad 还是Intern，Uber有统一流程，你必须从Uber网站 -> University Recruiting里提交。
提交简历给我们也没用。
10. 能否分享一下你对于ATG的个人感受？
ATG (as any company) is not perfect. There are definitely many areas for improvement. However a few things stand out as uniquely "cool" for me.
There's only lightweight processes and relatively little bureaucracy. Engineers spend most of our time doing the work. Decisions are often made quickly to move things along.
ATG is very much engineer driven - not PM or data driven. (there could be pros and cons here). Solid technical work counts A LOT.
We do work hard but there are no crazy on-call schedules - there's virtually nothing that requires us to wake up at nights and weekends.
This is a highly technical place full of geeky people. People will talk about tech with you, WAY more than gossip.
11. 如何申请？
本帖隐藏内容需要登录后才能查看。
- 16
2条回复