How to get your Shopify data ready for agentic AI

Sync Shopify orders, products, customers, and inventory with Airbyte so AI agents can answer commerce questions from a governed, read-only copy.

|
Nov 16, 2026

Summarize with AI:

Airbyte syncs Shopify orders, products, customers, and inventory into your warehouse, and the Airbyte context layer lets agents read and search Shopify records directly through the Airbyte MCP.

Commerce teams ask the same questions every day. What sold yesterday, which products are running low, how many orders were refunded, which discount drove the most revenue. An AI agent can answer them, and can go further by combining store data with ad spend, analytics, and payments. To do that reliably, it needs a complete copy of the store's data, not a series of calls against an API built for running a storefront.

This guide covers how Airbyte prepares Shopify data for agents: what it syncs, how it handles deletions and API limits, and how agents read the result.

Why Shopify data matters for agentic AI

A Shopify store records the full commercial story:

  • Orders and line items with prices, discounts, taxes, and fulfillment status.
  • Products, variants, and collections, the catalog an agent needs to talk about what you sell.
  • Customers and addresses, for segmentation and service.
  • Inventory items and levels by location.
  • Money movement: transactions, refunds, disputes, balance transactions, and tender transactions.
  • Marketing context: abandoned checkouts, discount codes, price rules, and customer journey summaries.

What agents can do with Shopify data

  1. Report daily performance. Summarize orders, revenue, refunds, and top products for any period from the warehouse copy.
  2. Watch inventory. Flag variants whose inventory levels are low at a given location, and relate them to recent sales velocity.
  3. Analyze discounts. Compare orders with and without specific discount codes or price rules to see what each promotion returned.
  4. Support customer service. Look up a customer's recent orders, fulfillments, and refunds to answer a ticket.
  5. Recover abandoned checkouts. List recent abandoned checkouts with product details for a follow-up workflow.

How Airbyte gets Shopify data ready

Authentication

On Airbyte Cloud, authenticate with OAuth 2.0. When you use OAuth, the connector works only in Airbyte Cloud or Airbyte Agents, because Airbyte's Shopify app accepts Admin API requests only from Airbyte's IP addresses. On self-managed Airbyte, create a Shopify custom app, grant only read_ scopes, and use its Admin API access token. API password authentication is being phased out on Airbyte Cloud.

APIs and sync modes

The connector reads through Shopify's REST API, GraphQL API, and GraphQL Bulk API, choosing per stream. It supports Full Refresh and Incremental syncs. Bulk operation limits set by Shopify apply to bulk streams.

Streams

The catalog includes:

  • Orders, Order Refunds, Order Risks, Order Agreements, Draft Orders, Abandoned Checkouts
  • Products, Product Variants, Product Images, Collections, Collection Products, Custom Collections, Smart Collections, Collects
  • Customers, Customer Address, Customer Journey Summary
  • Inventory Items, Inventory Levels, Locations
  • Fulfillments, Fulfillment Orders
  • Transactions, Tender Transactions, Balance Transactions, Disputes
  • Discount Codes, Discount Codes Sync, Price Rules
  • Metafields for many object types, plus Shop, Pages, Blogs, Articles, Countries, and Market Countries

Deletes

Deletions are captured for Articles, Blogs, Custom Collections, Orders, Pages, and Price Rules. The connector emits a record with the deleted ID and deleted_at, deleted_message, and deleted_description fields. For products, a dedicated Deleted Products stream reads product deletion events.

Known edge cases

  • Discount codes. Shopify's Bulk Operations API truncates nested codes at about 100 per discount without warning. Enable Discount Codes Sync if any discount has more than about 100 codes. Both streams share a schema, so you can deduplicate by id.
  • Market-driven shipping. Shopify is moving shipping configuration to Markets, with a rollout that started October 1, 2026. For migrated shops, the Countries stream keeps a frozen snapshot, and the Market Countries stream (which needs the read_markets scope) holds current configuration.

Controlling personal data

The Orders stream includes top-level copies of customer attributes such as customer_id, customer_email, and customer_default_address, populated when you enable Populate top-level customer fields in Orders. To give agents only the customer fields they need, deselect the nested customer object, select the specific customer_* fields, and configure hashing on them. The docs note that hashing customer_email alone does not remove the nested customer.email value while customer stays selected.

How agents reach Shopify data

Shopify is an agent connector in Airbyte Cloud:

  • Agent access (direct reads). Agents can list and get customers, orders, products, variants, inventory, locations, price rules, discount codes, collections, draft orders, fulfillments, refunds, transactions, and metafields. Direct reads use the credentials stored in Airbyte, need no sync, and are read-only. They count toward Shopify's API rate limits.
  • Search. When a connection syncs Shopify to an agent-connector destination, the selected streams can be indexed. The open-source Airbyte agent SDK lists semantic search for products, useful for questions like "do we sell anything for cold-weather running."
  • Warehouse SQL. Synced to BigQuery or Snowflake, Shopify history is available to agents through read-only SQL.

Agents connect through the Airbyte MCP at https://mcp.airbyte.com/mcp. Direct reads and searches are free up to 100,000 combined per organization per month, and the context layer is available only in Airbyte Cloud.

What this unlocks

  • Daily commerce reporting from a complete copy of orders and refunds.
  • Inventory and catalog questions answered without storefront API calls.
  • Deletions tracked for key objects, so counts stay accurate.
  • Customer data limited or hashed before agents see it.
  • Store data joined with ads, analytics, and payments in the warehouse.

The practical difference for an agent is that its questions stop depending on what one API call can return. A question like "which products sold best last quarter among repeat customers" needs orders, line items, and customer history together, which is a query against the synced tables rather than a series of storefront requests.

Getting started

  1. Create a Shopify source in Airbyte Cloud and connect with OAuth.
  2. Select order, product, customer, and inventory streams, and enable Discount Codes Sync if needed.
  3. Configure customer field selection and hashing on Orders.
  4. Sync to BigQuery or Snowflake, then connect your agent to the Airbyte MCP.

Try Airbyte free or talk to sales about commerce data for agents.

Frequently asked questions

Which Shopify authentication should I use with Airbyte?

Use OAuth 2.0 on Airbyte Cloud. With OAuth, the connector only runs in Airbyte Cloud or Airbyte Agents. On self-managed Airbyte, create a Shopify custom app with read_ scopes and use its Admin API access token.

Does Airbyte capture deleted Shopify records?

Yes, for some streams. Deletions are captured for Articles, Blogs, Custom Collections, Orders, Pages, and Price Rules, and a dedicated Deleted Products stream returns product deletion events.

Why are some discount codes missing from my Shopify sync?

Shopify's Bulk Operations API truncates nested discount codes at roughly 100 per discount. Enable the Discount Codes Sync stream, which pages through every code, if any discount has more than about 100 redeem codes.

Can I limit customer personal data that agents see?

Yes. In the Orders stream you can deselect the nested customer object, select only the top-level customer fields you need, and configure hashing on them. Hashing customer_email alone does not remove the nested value while the customer object stays selected.

Can an agent edit products or orders through the Airbyte context layer?

No. Direct reads and searches through the context layer are read-only. Agents can list, get, and search Shopify records, but cannot create, update, or delete them through these tools.

Integrate with 700+ apps using Airbyte

Move data from 700+ sources into warehouses, lakes, and beyond. Set up pipelines in minutes with pre-built connectors and the Connector Builder.