Fashion. Style. Elegance. These are words that are easy to appreciate and understand for a human, but hard for algorithms and computers. Recommending fashion – yes, that is the space that we work in. We blend a variety of disciplines (such as recommender systems, data mining, NLP, big data, semantic web, graph stores and cloud computing) to turn an intractable problem into a hard problem. For example: just because someone bought a purple dress does not mean that purple is the right color for them for all seasons; neither does it mean that you should recommend purple accessories. Just when someone buys a shirt, which tie in Amazon’s vast selection would it make most sense to pair with it? While there are other recommendation problems out there, none of them capture the scale of the problem (think size times color), the rapidly changing inventory (each season changes what’s cool) along with unique problems of fit, fabric and finish.
So how do we do it? We are a full stack team that own components all the way from dataset generation, high performance service oriented architecture to building great UIs that can surface recommendations in a pleasing and aesthetically appealing way. We work with a talented group of UX designers, data scientists, machine learning experts and UI developers. As for techniques, we experiment with collaborative filtering, matrix factorization and have recently started using deep learning. You don’t need to be fashionista or a shopaholic. But, if you know what peplum means and colorblocking is more than just the two words put together, you are already ahead of the curve (or perhaps you watch Project Runway…. which also works!). If you want to be on the cutting edge of personalization in ecommerce retail and reaching customers in a unique and valuable space, this is the team to be on! We have the mandate and ability to effect big changes, we just need the right person to begin.
As a Software Development Engineer, you will own and build infrastructure that accesses terabyte of data to produce and deliver datasets with low latency and high reliability. The goal is to assemble features and services quickly that can make a huge impact on the customer experience. What recommendations data could you use to make search more personalized? How can we use product similarity and sales data to augment and enrich our recommendations? Software Engineers at Amazon are more than just order takers; they see a problem and leverage innovative technology to address it. We're looking for people who innovate, love solving hard problems, and never take "no" for an answer. You will be able to build systems that will impact millions of customers and create multi-million dollar revenue opportunities, ship in just a couple of weeks and instantly measure the dollar value of what you have developed. And be able to go home and show your friends and family the impact of what you are doing!
- Analyze and extract relevant information from large amounts of Amazon's historical business data to help automate and optimize key features and processes.
- Work closely with stakeholders to optimize various business operations
- Developing recommendation algorithms to power features on Amazon
- Inferring customer intent to suggest relevant items, categories or brands to all customers
- Finding powerful product relationships,
- Reducing latency to make the shopping experience blazingly fast