TechPrufer
Get started
Shopify
Shopify logoShopify · Feature trail

Tax Engine Componentization

2 posts · Jul 2019 – Feb 2023

What this trail covers

Shopify has evolved its tax engine by developing and enhancing the Tax Insights feature. This involves complex data modeling, creating new data models, handling dynamically changing tax laws and economic nexus thresholds, and publishing these insights to a key-value store for end-user consumption. The architecture leverages data extraction into a data warehouse (GCS), processing with Spark jobs, and serving data via Bigtable. This enables merchants to stay informed about their sales tax obligations.

All Shopify trails
  1. The Complex Data Models Behind Shopify's Tax Insights Feature - ShopifyThis post details the development of the Tax Insights feature, focusing on the data architecture and modeling required for tax compliance. It describes the process of understanding complex tax laws, designing and prototyping data models using SQL and BigQuery, and implementing these models using Spark (Starscream/Seamster) to calculate economic nexus and tax liabilities for merchants. The final data is published to Bigtable for frontend display and also made available for analytics.
  2. Componentizing Shopify’s Tax Engine - ShopifyThis post details the componentization of Shopify's tax engine. The team educated themselves on the tax domain, identified and refactored existing tax logic spread across the codebase, defined a clear interface with `TaxesRequestSchema` and `TaxesResponseSchema`, moved all tax-related source files into a new tax component, and implemented GitHub bot triggers to enforce boundaries. They updated testing suites with unit and integration tests, and released the component through an experimental code path that ran old and new calculations in parallel for comparison, followed by a phased rollout to 100% of shops.