Key Responsibilities:

Own and develop advanced substitutability analysis frameworks combining text-based and visual matching capabilities
Drive technical improvements to product matching models to enhance accuracy beyond current 79% in structured categories
Design category-specific matching criteria, particularly for complex categories like fashion where accuracy is currently at 20%
Develop and implement advanced image matching techniques including pattern recognition, style segmentation, and texture analysis
Create performance measurement frameworks to evaluate product matching accuracy across different product categories
Partner with multiple data and analytics teams to integrate various data signals
Provide technical expertise in scaling substitutability analysis across 2000 different product types in multiple markets

Technical Requirements:

Deep expertise in developing hierarchical matching systems
Strong background in image processing and visual similarity algorithms
Experience with large-scale data analysis and model performance optimization
Ability to work with multiple data sources and complex matching criteria



Key job responsibilities
Success Metrics:

Drive improvement in substitutability accuracy to >70% across all categories
Reduce manual analysis time for product matching identification
Successfully implement enhanced visual matching capabilities
Create scalable solutions for multi-market implementation


A day in the life
Design, develop, implement, test, document, and operate large-scale, high-volume, high-performance data structures for business intelligence analytics. Implement data structures using best practices in data modeling, ETL/ELT processes, SQL, Oracle, and OLAP technologies. Provide on-line reporting and analysis using OBIEE business intelligence tools and a logical abstraction layer against large, multi-dimensional datasets and multiple sources. Gather business and functional requirements and translate these requirements into robust, scalable, operable solutions that work well within the overall data architecture. Analyze source data systems and drive best practices in source teams. Participate in the full development life cycle, end-to-end, from design, implementation and testing, to documentation, delivery, support, and maintenance. Produce comprehensive, usable dataset documentation and metadata. Evaluate and make decisions around dataset implementations designed and proposed by peer data engineers. Evaluate and make decisions around the use of new or existing software products and tools. Mentor junior Business Research Analysts.

About the team
The RBS-Availability program includes Selection Addition (where new Head-Selections are added based on gaps identified by Selection Monitoring-SM), Buyability (ensuring new HS additions are buyable and recovering established ASINs that became non-buyable), SoROOS (rectify defects for sourceble out-of-stock ASINs ) Glance View Speed (offering ASINs with the best promise speed based on Store/Channel/FC level nuances), Emerging MPs, ASIN Productivity (To have every ASINS actual contribution profit to meet or exceed the estimate). The North-Star of the Availability program is to "Ensure all customer-relevant (HS) ASINs are available in Amazon Stores with guaranteed delivery promise at an optimal speed." To achieve this, we collaborate with SM, SCOT, Retail Selection, Category, and US-ACES to identify overall opportunities, defect drivers, and ingress across forecasting, sourcing, procurability, and availability systems, fixing them through UDE/Tech-based solutions.