Address Intelligence and Experience (AIX) team is looking for an Applied Scientist based out of Hyderabad, India. As part of the Amazon Delivery Technology group, this team is responsible for ensuring best-in-class delivery experience for customers who shop on Amazon. The AIX team does this by learning all we can about every possible delivery location on the planet and using these inputs to drive accurate, efficient and low cost delivery to our customers. We aim to make every address printed on an Amazon shipping label accurate and deliverable, while ensuring we have additional intelligence such as access codes, location photos, geocodes, business hours, and customer delivery preferences. To achieve this worldwide, we focus on building comprehensive address data models for the regions we operate in, and on developing sophisticated ML based software that recognizes and validates customer addresses, learns from historical data as well as through crowd sourced intelligence. A unique challenge for our team is that while all our platforms and technology have to be global and scalable, our solutions are always highly customized for each region given that addresses are structured and managed very differently across countries.
The Applied Scientist will support Address Intelligence and Experience (AIX) teams’ growth by designing, developing, evaluating, deploying and updating of data-driven models and analytical solutions for machine learning to solve practical problems that our Customers and systems face today. This could be the drivers who are delivering packages, or Customers who are entering or providing us address data, or systems that match address to a reference database. You will work with software developers and other teams with a mandate to
· Use (and not limited to) Machine Learning Techniques to detect anomalies in address data to reduce delivery defects.
· Use Natural Language Processing techniques to detect addresses or related metadata from text.
· Use (and not limited to) statistical models to improve address resolution and model real world addresses (identify buildings, communities, campuses etc.).
· Implement statistical models for recommending the right questions to collect address data at the right time considering context and behavior of a customer.
You will also demonstrate superior verbal and written communication and presentation skills, ability to convey rigorous mathematical concepts and considerations to non-experts. We are looking for a team-player to play a significant part in defining our team efforts. The successful candidate will be a self-starter, comfortable with ambiguity and be able to create and maintain efficient & automated processes. He / She will have to work alongside with stakeholders in the organization to make it happen. You are analytical and creative, and you don’t quit. This is a role with high visibility to senior leadership and with high opportunity for impact for those willing to roll up their sleeves and dive deep to achieve results.