Depop operates a mobile-first social commerce platform for buying and selling preloved fashion, now owned by Etsy. The tech stack reveals a data-engineering-heavy organization: Kafka, Databricks, dbt, and Apache Airflow sit alongside ML infrastructure (PyTorch, TensorFlow, Transformers), while active projects center on data contracts, ranking models, and multi-touch attribution. Hiring velocity is accelerating across engineering and data roles, with pain points clustering around search relevance, payment system reliability, and real-time personalization—core challenges for a high-volume marketplace managing both seller trust and buyer discovery.
Depop is a circular fashion marketplace where users buy, sell, and discover secondhand clothing. The platform has 43M registered users globally and operates as a standalone subsidiary of Etsy since 2021. The company is headquartered in London with offices in New York and approximately 400 employees. The business runs on a mobile-native model (Swift, iOS, Kotlin, Android) with backend services on AWS, PostgreSQL, and DynamoDB. Core infrastructure includes Kafka for event streaming, Databricks and dbt for data transformation, and machine-learning systems for ranking and content moderation. Search relevance, payment system resilience, and real-time personalization are ongoing operational priorities.
Depop's stack spans mobile (Swift, iOS, Kotlin, Android), compute (AWS, Kubernetes, Terraform), data (PostgreSQL, DynamoDB, Kafka, Databricks, dbt, Apache Airflow), ML (PyTorch, TensorFlow, Transformers), monitoring (Datadog, Looker), and payments (Stripe). Scala and Java handle backend services.
Current projects include ranking models for the app, data contracts, data observability systems, multi-touch attribution, central data models for ML and search, measurement roadmap development, and in-app support entry points. Payment system reliability and search relevance are active operational challenges.
Other companies in the same industry, closest in size