Software Development Engineer - Display Advertising, Targeting and Prediction Systems
Amazon.com's Display Advertising team builds and manages systems with extreme high performance and availability. We serve and respond to hundreds of billions of requests annually, and have ambitions to grow that number several orders of magnitude, while maintaining response latencies in the milliseconds and meeting strict SLA requirements. We delight in data, and are constantly trying to enrich our models and make more informed decisions using every bit and byte we have. Our current systems serve our current needs, but we are growing a new business within Amazon.com, and are looking for the sort of hands on leaders and developers who can design and build the systems that will take us into the future.
About you:
You're an engineer looking for a career where you'll be able to build, to deliver, and to impress. You look at problems holistically, and thrive on the intricate complexity of designing feedback loops and ecosystems. You want to work on projects where you are implementing solutions to real problems that require creative solutions and deep understanding of the problem space. You challenge yourself and others to constantly come up with better solutions. You're a thought leader, but you don't just know how to solve the problem, you prove it by going and building the solution. You want to create value and impact the direction of entire industries.
About us together:
We're going to change the way that the advertising world measures, plans, and buys. Along the way, 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. We'll make tough decisions, but we'll all understand why. We'll be the dream team.
Some problem spaces we'll be working on:
TARGETING - we have a wealth of data already about the customers we interact with; what they've been shopping for, what they've recently purchased, etc. This data can be used to help us decide what else they are likely to be interested in and to respond to. Accessing this data and making decisions across large numbers of potential responses in a matter of milliseconds is a tough problem, and requires complex high scale engineering both online and offline. You may specialize in one area or another, but will need to consider the problem holistically.
PREDICTIVE MODELING - there's an almost overwhelming amount of data available, and we need to build the best models to predict which signals are likely to have an impact and which aren't. We're building platforms to automate the generation, validation, and publishing of models that incorporate all available data to predict any measurable signal. Youll need to be both an engineer and scientist youll build the platforms to test and measure your hypotheses, iterate on your designs, and find innovative signals and solutions that fundamentally shift the effectiveness of our systems.
INVENTORY MANAGEMENT - advertisers spend a lot of money to develop their ads, and want a guarantee that those ads will actually get shown. We need to be able to make those guarantees and optimize future inventory allocation across thousands of independent dimensions in near to real time. You'll need to design a system that will be able to manage the complexity of thousands of semi-independent overlapping feature dimensions to predict future inventory volume on any combination of those dimensions, and optimize the allocation of that predicted inventory, all while providing the ability to responding to inquiries and honor reservation requests. Beyond this, the right solution should be able to explicitly calculate the full system cost and risk trade-offs of any such guarantee, and determine what price if any we should set on the deal.
LIFT ANALYSIS - simply measuring clicks and conversions is not enough any advertiser will tell you that the goal of advertising is to change the behaviour of their potential customers. Knowing that a customer purchased a product after seeing an ad does not necessarily imply that the customer's behaviour was changed by seeing the ad. Teasing apart causation from correlation is difficult at its best, and seemingly impossible at worst, but we know that seemingly impossible problems are just opportunities awaiting creative solutions. We seek to solve this for our advertisers, to help them measure multiple signals, and to work with and educate them on where their spend is producing desired results, and where it's not.
Questions?:
You may already know if you're a fit, but perhaps you're worried about technology and experience requirements? Don't be - we're looking for smart, proven, engineers; if you're the right candidate, we're flexible. You're still worried? Here are some things that we'd like to see:
* online distributed systems and platform infrastructure design and implementation
* experience in the advertising, algorithmic search, or finance technology industries
* experience implementing servers using Java or C++ in a Linux/Unix environment
* delivered a low-latency, consumer-internet-scale web services operating in a 24x7 environment
* delivered ad targeting and serving infrastructure
* previously held a technical leadership role for several complete large-scale projects
* experience with Agile (SCRUM, RUP, XP), OO modeling, web services, UNIX, middleware, database and data mining systems
* experience mentoring and training the engineering community on complex technical issues
* sharp analytical abilities and proven design skills
* track record of delivery
* 5+ years of industry experience
Please send resumes directly to nkrishna@amazon.com![]()
Amazon.com's Display Advertising team builds and manages systems with extreme high performance and availability. We serve and respond to hundreds of billions of requests annually, and have ambitions to grow that number several orders of magnitude, while maintaining response latencies in the milliseconds and meeting strict SLA requirements. We delight in data, and are constantly trying to enrich our models and make more informed decisions using every bit and byte we have. Our current systems serve our current needs, but we are growing a new business within Amazon.com, and are looking for the sort of hands on leaders and developers who can design and build the systems that will take us into the future.
About you:
You're an engineer looking for a career where you'll be able to build, to deliver, and to impress. You look at problems holistically, and thrive on the intricate complexity of designing feedback loops and ecosystems. You want to work on projects where you are implementing solutions to real problems that require creative solutions and deep understanding of the problem space. You challenge yourself and others to constantly come up with better solutions. You're a thought leader, but you don't just know how to solve the problem, you prove it by going and building the solution. You want to create value and impact the direction of entire industries.
About us together:
We're going to change the way that the advertising world measures, plans, and buys. Along the way, 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. We'll make tough decisions, but we'll all understand why. We'll be the dream team.
Some problem spaces we'll be working on:
TARGETING - we have a wealth of data already about the customers we interact with; what they've been shopping for, what they've recently purchased, etc. This data can be used to help us decide what else they are likely to be interested in and to respond to. Accessing this data and making decisions across large numbers of potential responses in a matter of milliseconds is a tough problem, and requires complex high scale engineering both online and offline. You may specialize in one area or another, but will need to consider the problem holistically.
PREDICTIVE MODELING - there's an almost overwhelming amount of data available, and we need to build the best models to predict which signals are likely to have an impact and which aren't. We're building platforms to automate the generation, validation, and publishing of models that incorporate all available data to predict any measurable signal. Youll need to be both an engineer and scientist youll build the platforms to test and measure your hypotheses, iterate on your designs, and find innovative signals and solutions that fundamentally shift the effectiveness of our systems.
INVENTORY MANAGEMENT - advertisers spend a lot of money to develop their ads, and want a guarantee that those ads will actually get shown. We need to be able to make those guarantees and optimize future inventory allocation across thousands of independent dimensions in near to real time. You'll need to design a system that will be able to manage the complexity of thousands of semi-independent overlapping feature dimensions to predict future inventory volume on any combination of those dimensions, and optimize the allocation of that predicted inventory, all while providing the ability to responding to inquiries and honor reservation requests. Beyond this, the right solution should be able to explicitly calculate the full system cost and risk trade-offs of any such guarantee, and determine what price if any we should set on the deal.
LIFT ANALYSIS - simply measuring clicks and conversions is not enough any advertiser will tell you that the goal of advertising is to change the behaviour of their potential customers. Knowing that a customer purchased a product after seeing an ad does not necessarily imply that the customer's behaviour was changed by seeing the ad. Teasing apart causation from correlation is difficult at its best, and seemingly impossible at worst, but we know that seemingly impossible problems are just opportunities awaiting creative solutions. We seek to solve this for our advertisers, to help them measure multiple signals, and to work with and educate them on where their spend is producing desired results, and where it's not.
Questions?:
You may already know if you're a fit, but perhaps you're worried about technology and experience requirements? Don't be - we're looking for smart, proven, engineers; if you're the right candidate, we're flexible. You're still worried? Here are some things that we'd like to see:
* online distributed systems and platform infrastructure design and implementation
* experience in the advertising, algorithmic search, or finance technology industries
* experience implementing servers using Java or C++ in a Linux/Unix environment
* delivered a low-latency, consumer-internet-scale web services operating in a 24x7 environment
* delivered ad targeting and serving infrastructure
* previously held a technical leadership role for several complete large-scale projects
* experience with Agile (SCRUM, RUP, XP), OO modeling, web services, UNIX, middleware, database and data mining systems
* experience mentoring and training the engineering community on complex technical issues
* sharp analytical abilities and proven design skills
* track record of delivery
* 5+ years of industry experience
Please send resumes directly to nkrishna@amazon.com
Source: Joel On Software