Augmented commerce: Machine learning at Shopify

Summary

Shopify frames commerce as a complex, continuous optimization problem in which merchants have different definitions of success and need support across the business lifecycle. Its machine-learning work uses a historical dataset covering millions of businesses to optimize product metadata, transaction fraud assessment, merchant assistance, product discovery, capital decisions, behavioral modeling, and search query rewriting. The systems include fine-tuned Qwen multimodal models, LLaMa models and MCPs for Sidekick, Nomic embeddings for billions of products, a tabular transformer for forecasting merchant GMV, and experiments with HSTU sequence models. To support this work, Shopify partners with GCP, Nebius, CentML, and Toloka, while providing access to GPUs, data services, frontier models, and internal tooling; the post also presents these efforts as part of its hiring pitch.

Context

Shopify describes commerce as a wide, continuously changing optimization problem: merchants must create, present, and sell products while understanding their environment and controlling their businesses. The company aims to help merchants succeed despite each having different objectives, and says its machine-learning investments address problems across the commerce lifecycle.

Approach / What changed

Shopify partitions the broader optimization problem into finite machine-learning systems, using a historical dataset from millions of businesses. Its approach combines specialized models and services for product metadata, fraud, merchant assistance, embeddings, GMV forecasting, behavior modeling, and query rewriting with GPU infrastructure, cloud-provider flexibility, data services, frontier models, and internal tooling.

Takeaways

• Fine-tuned Qwen multimodal models classify and enrich metadata for every product uploaded to Shopify, requiring hundreds of millions of inferences each day.
• Nomic embeddings provide vector representations for many billions of catalog products and support product discovery systems such as search and recommendations.
• Shopify uses a tabular transformer-based model to forecast merchant GMV as part of risk assessment for providing merchants with access to capital.

Related reading