Product Data and Matching Engineer

Role overview Develop the data quality and product-matching layer for an AI-powered shopping search service. Listings for the same item can vary widely by title, seller, and version; your work will connect matching offers while distinguishing meaningful differences such as condition and market.

Responsibilities - Match listings across stores using brand, model, variant, and identifiers such as GTIN or MPN. - Apply embeddings and language models where useful, supported by rules and evaluation data. - Classify offers by condition, seller type, and country to make results easier to filter. - Build sets of real search queries and track matching precision and recall over time. - Use quality findings to improve extraction plans, search behavior, and offer ranking.

Requirements - At least three years in data engineering, search, or applied machine learning with messy real-world data. - Strong Python and SQL skills. - Experience creating evaluation sets and measuring system quality. - Practical experience using language model APIs or embeddings. - Clear written English.

Nice to have - Experience with ecommerce catalogs, product taxonomies, or entity resolution. - Familiarity with vector search or ranking systems.