Software Development Engineer

Company: Amazon.com
Location: Seattle, WA, USA
Web: www.amazon.com

_Contact_:
Email to nkrishna@amazon.com , refer to Job number: 110739

Job Description:
Does processing semi-structured or un-structured data into structured information interest you? Do you like taking on new challenges and exploring new frontiers in pursuit of combing information from any source?

The Fact Extraction (FX) team specializes in processing semi-structured merchant contributions to extract meaningful structured information to augment and correct Amazon's Item Catalog. Our goal is to enhance customer experience by improving accuracy and consistency of key product attributes (like brand), improving browse and search experience, and providing clean and consistent information for product variations. We have developed our own extraction framework that allows us to combine information retrieval and knowledge representation techniques to conquer complex extraction problems while processing tens of millions of merchant contributions daily. We have also built and continue to enhance our asynchronous analysis and reporting framework that monitors the quality of our extraction results. We are continually expanding our reach to larger sections of the catalog and in the process tackling interesting and challenging information retrieval problems particular to a given section of the catalog; e.g. analyzing data to discover and model knowledge; designing and implementing information retrieval techniques; enhancing the framework, and measuring quality and customer impact.

This is a great time to join our team as we are poised to tackle complex information retrieval challenges, expand our infrastructure and further enhance customer experience on Amazon. Our vision is to continually improve the accuracy and breadth of our extraction framework by using techniques such as machine based knowledge discovery, NLP based extraction techniques, and augment information by scraping sources like the internet, to name a few.

As a developer on this team, you will have the opportunity to:

• Dive deep into the catalog data, understand different functional areas, and use your creativity to come up with extraction techniques that improve the catalog quality issues on hand.
• Identify areas of improvement in our frameworks, tools, processes and strive to make them better. Evaluate our success metrics and evolve our reporting systems.
• Work with the category owners within Amazon to understand data quality issues, functional specifics of data and contributions.

• Writing high quality code and participating in code reviews.
• Participating in team meetings, stand-ups, and architecture/design discussions.
• Following Amazon's ownership model, own extraction rules deployed in production, and follow reporting metrics and continually come up with ways to improve them.
• Working on improvements to infrastructure; FX extraction framework or reporting system.
• Sharing the operational load by being on a regular on-call rotation.

• Results oriented person with a delivery focus.
• Strong coding skills in Java coupled with strong base in object-oriented design and development.
• Excellent verbal and written communication skills.
• 4-6 years experience in Software development.
• Ability to handle multiple competing priorities in a fast-paced environment.
• Experience in information retrieval, knowledge representation or computational linguistics is a plus.
• Fluency in written Chinese/Japanese/German is a plus.