Skip to main content

Software Development Engineer-Machine Learning


Job Description

Amazon's Product Advertising program handles the placement of hundreds of millions of ad impressions every day on Amazon.com and other websites (See http://www.amazonservices.com/productAds/index.html). The optimization team plays a critical role in improving ad targeting and optimizing the revenue we realize from displaying ads.

We apply data mining and machine learning on terabytes of data to build intelligent systems addressing key problems in the online advertising space: when we show an ad, how likely is it to get clicked? What are the ads which are most relevant given the page content? Does a website have objectionable content?
We have a tight cycle of data mining, prototyping, and testing in production which allows us to make iterative improvements quickly.

Our business is rapidly expanding into new areas and bringing in new, challenging problems. We're seeking a skilled and innovative software engineer with aptitude in applied mathematics to help us build systems that can target ads better and realize higher revenue in the process.

About you:
You're an engineer looking for a career where you'll be able to build, to deliver, and to impress. You enjoy getting computers to make intelligent decisions in the face of real-world noisy data. You think statistics is a cool subject. You are excited at the opportunity of applying cutting-edge machine learning techniques to problems that significantly impact Amazon’s business. You can read a technical paper, understand and adapt an idea, build a prototype, release production code, and interpret the results rigorously.
You challenge yourself and others to constantly come up with better solutions. You're a thought leader and you demonstrate this by building solutions, not just by having ideas. You develop strong working relationships with others and want to work in a collaborative team environment. 

About us together:
We're going to design and deliver the next generation of ad targeting and ad optimization systems.  We're going to face seemingly impossible problems. We're going to argue about how to solve them, and we'll work together to find a solution that is superior to each of the proposals we came in with. At the end of the day, we will deliver something of great value to the Product Ads business and to Amazon, and we'll have fun and improve our skills along the way.

A few problem spaces we'll be working on:

Click-through-rate estimation – We show tens of millions of ads to tens of millions of users on tens of millions of pages daily. Ad clicks are rate events, yet it is important to be able to predict the probability of click of an ad accurately for any given display situation. How do you build a system that can make good predictions? How do you measure its accuracy? How do you improve it? And how do you ensure it scales to handle billions of predictions per day, in real-time?

Managing advertiser budgets – Our advertisers provide a daily budget which sometimes runs out. How can we “stretch” their budgets intelligently, so they can be exposed to traffic throughout the day? How can we expose merchants which are running low on budgets to only the most profitable opportunities? How do we do better budget forecasting to help merchants plan their budgets better?


What we are looking for:

A smart, proven engineer with a solid math and statistics background with a love for working with large data and building scalable intelligent systems.

Basic Qualifications

  • Bachelor’s Degree in Computer Science, Math or related field
  • 3+ years professional experience in software development
  • Strong Computer Science fundamentals in data structures, algorithm design, problem solving, and complexity analysis
  • Mastery of at least one modern programming language such as Java or C/C++ and at least one scripting language such as Perl, Python, or Ruby

Preferred Qualifications

  • MS or PhD in Computer Science, Math, or related field
  • Experience building large-scale data mining systems
  • Experience with algorithm development involving real-world noisy data
  • Experience with application of machine-learning techniques
  • Experience with statistical software such as R, SAS, or weka
  • Experience delivering low-latency, consumer-internet-scale web services operating in a 24x7 environment
  • Experience mentoring and training others on complex technical issues
  • Proven ability to take a project from scoping requirements through actual launch of the project
  • Solid coding practices including good design documentation, unit testing, peer code reviews, and a preference for agile methods
  • Sharp analytical abilities, proven design skills, excellent communication skills
  • Strong sense of ownership, urgency, and drive, and a track record of delivery
  • Passion for building new products in a fast paced, team-oriented environment.
  • 5+ years of industry experience

Comments

Popular posts from this blog

OWASP Top 10 Threats and Mitigations Exam - Single Select

Last updated 4 Aug 11 Course Title: OWASP Top 10 Threats and Mitigation Exam Questions - Single Select 1) Which of the following consequences is most likely to occur due to an injection attack? Spoofing Cross-site request forgery Denial of service   Correct Insecure direct object references 2) Your application is created using a language that does not support a clear distinction between code and data. Which vulnerability is most likely to occur in your application? Injection   Correct Insecure direct object references Failure to restrict URL access Insufficient transport layer protection 3) Which of the following scenarios is most likely to cause an injection attack? Unvalidated input is embedded in an instruction stream.   Correct Unvalidated input can be distinguished from valid instructions. A Web application does not validate a client’s access to a resource. A Web action performs an operation on behalf of the user without checking a shared sec

CKA Simulator Kubernetes 1.22

  https://killer.sh Pre Setup Once you've gained access to your terminal it might be wise to spend ~1 minute to setup your environment. You could set these: alias k = kubectl                         # will already be pre-configured export do = "--dry-run=client -o yaml"     # k get pod x $do export now = "--force --grace-period 0"   # k delete pod x $now Vim To make vim use 2 spaces for a tab edit ~/.vimrc to contain: set tabstop=2 set expandtab set shiftwidth=2 More setup suggestions are in the tips section .     Question 1 | Contexts Task weight: 1%   You have access to multiple clusters from your main terminal through kubectl contexts. Write all those context names into /opt/course/1/contexts . Next write a command to display the current context into /opt/course/1/context_default_kubectl.sh , the command should use kubectl . Finally write a second command doing the same thing into /opt/course/1/context_default_no_kubectl.sh , but without the use of k

标 题: 关于Daniel Guo 律师

发信人: q123452017 (水天一色), 信区: I140 标  题: 关于Daniel Guo 律师 关键字: Daniel Guo 发信站: BBS 未名空间站 (Thu Apr 26 02:11:35 2018, 美东) 这些是lz根据亲身经历在 Immigration版上发的帖以及一些关于Daniel Guo 律师的回 帖,希望大家不要被一些马甲帖广告帖所骗,慎重考虑选择律师。 WG 和Guo两家律师对比 1. fully refund的合约上的区别 wegreened家是case不过只要第二次没有file就可以fully refund。郭家是要两次case 没过才给refund,而且只要第二次pl draft好律师就可以不退任何律师费。 2. 回信速度 wegreened家一般24小时内回信。郭律师是在可以快速回复的时候才回复很快,对于需 要时间回复或者是不愿意给出确切答复的时候就回复的比较慢。 比如:lz问过郭律师他们律所在nsc区域最近eb1a的通过率,大家也知道nsc现在杀手如 云,但是郭律师过了两天只回复说让秘书update最近的case然后去网页上查,但是上面 并没有写明tsc还是nsc。 lz还问过郭律师关于准备ps (他要求的文件)的一些问题,模版上有的东西不是很清 楚,但是他一般就是把模版上的东西再copy一遍发过来。 3. 材料区别 (推荐信) 因为我只收到郭律师写的推荐信,所以可以比下两家推荐信 wegreened家推荐信写的比较长,而且每封推荐信会用不同的语气和风格,会包含lz写 的research summary里面的某个方面 郭家四封推荐信都是一个格式,一种语气,连地址,信的称呼都是一样的,怎么看四封 推荐信都是同一个人写出来的。套路基本都是第一段目的,第二段介绍推荐人,第三段 某篇或几篇文章的abstract,最后结论 4. 前期材料准备 wegreened家要按照他们的模版准备一个十几页的research summary。 郭律师在签约之前说的是只需要准备五页左右的summary,但是在lz签完约收到推荐信 ,郭律师又发来一个很长的ps要lz自己填,而且和pl的格式基本差不多。 总结下来,申请自己上心最重要。但是如果选律师,lz更倾向于wegreened,