AI-powered ecommerce search and product discovery platform
Constructor builds an AI-first search and discovery platform purpose-built for ecommerce conversion optimization. The stack reveals a ML-heavy engineering org: PyTorch, Transformers, Spark SQL, and Databricks sit alongside real-time infrastructure (Kinesis, Elasticsearch, ClickHouse), and the hiring mix (78 engineers, 37 support, 21 product) reflects a product-led, engineering-intensive scaling phase. Active projects on query interpretation strategy and attribute enrichment suggest the company is deepening personalization logic, while pain points around integration robustness and onboarding indicate operational friction as adoption grows.
Notable leadership hires: Product Marketing Director, Account Director
Constructor develops a commerce search and discovery platform designed to maximize ecommerce metrics including conversion rate, revenue, and profit. The product surfaces across search, recommendations, collections, AI shopping assistants, and guided selling, powered by natural language processing, machine learning, and generative AI. The company serves enterprise retailers, distributors, and marketplaces globally. Founded in 2015 and headquartered in San Francisco, Constructor operates with 201–500 employees and is expanding rapidly across 25+ countries, with notable hiring concentration in Europe, the Americas, and Australia.
PyTorch, Transformers, and Apache Spark form the ML core. Real-time serving relies on Elasticsearch, ClickHouse, and Databricks, with Ray for distributed compute and FastAPI for serving layers.
Active projects include improving search and discovery experiences, attribute enrichment product strategy, query interpretation strategy, integration health monitoring, and scaling customer success teams across EMEA.
Other companies in the same industry, closest in size