Introduction

Intershop Copilot for Buyers is an AI-powered conversational buying and service assistant embedded in the Intershop storefront. It combines a buyer-facing PWA experience with AI-based conversation orchestration, commerce services, search, and retrieval capabilities, and configurable tools. Depending on the configured release and integrations, the Copilot can support product discovery, product questions, buyer-specific information, commerce operations, knowledge retrieval, and selected storefront actions.

This concept is intended for implementation partners, solution architects, developers, product managers, and technical administrators.

This concept provides a functional and architectural overview of Copilot for Buyers. It does not replace the installation guides, implementation cookbook, security concept, release documentation, or customer-specific solution design.

References

• Guide - Copilot for Buyers Starting Guide

• Concept - Security for Intershop Copilot for Buyers

• Azure AI Search indexers (learn.microsoft.com)

• Azure Blob Storage indexing with Azure AI Search (learn.microsoft.com)

• Chunk documents for vector search (learn.microsoft.com)

• Integrated vectorization in Azure AI Search (learn.microsoft.com)

• Azure Blob Storage documentation | Azure Storage Explorer (learn.microsoft.com)

• Intershop Copilot for Buyers in the PWA (github.com/intershop)

• Cookbook - Intershop Copilot for Buyers

Product Overview

Business Purpose

Intershop Copilot for Buyers is a conversational buying and service assistant integrated into an Intershop storefront. It helps buyers find and understand products, obtain answers from approved knowledge sources, access buyer-specific information, and complete supported commerce or storefront interactions.

The Copilot combines natural-language interaction with configured commerce, search, knowledge, and business services. The Copilot makes common buying and service tasks easier by allowing buyers to express their needs conversationally instead of navigating every process manually.

Product Boundaries

Copilot for Buyers is a conversational control and orchestration layer. It interprets buyer requests, retrieves information from configured services, invokes approved tools, and prepares a response or supported action.

It does not replace the systems that own the underlying data or business processes.

The IAB Chatflow or Agentflow interprets requests and coordinates the configured services. It combines retrieved information and prepares the response, but the connected systems remain responsible for their own data, permissions, and business rules.

Main components

The relationships between these components are shown in the Architecture Overview.

Capabilities

Intershop Copilot for Buyers supports conversational buying and service scenarios in an Intershop storefront. Depending on the configured flow and connected services, buyers can search for products, understand product information, retrieve answers from approved knowledge sources, access permitted buyer-specific information, and initiate supported storefront or commerce actions.

The following capability areas describe the current product scope at a conceptual level. Availability depends on the Copilot for Buyers release, the PWA and ICM versions, the configured IAB flow, connected services, and the buyer’s permissions.

The capabilities available in a Copilot for Buyers deployment depend on the selected Chatflow or Agentflow, configured tools, connected services, authentication context, PWA integration, and product release. The capabilities below describe supported or configurable capability areas, not a guarantee that every deployment provides every function.

Commerce assistance

Commerce assistance uses connected commerce services and the configured commerce-search provider to help buyers discover products and complete supported buying tasks.

Product discovery

The Copilot can help buyers find products using natural-language requests. A buyer can describe a product, requirement, or intended use instead of entering only a product identifier or predefined search phrase.

Product discovery can use:

• Product data from ICM

• The configured commerce-search service

• Product attributes and categories

• Buyer-provided requirements

• Customer-specific search or recommendation logic

The available search behavior depends on the configured commerce-search provider and the implementation of the selected flow. Commerce search remains responsible for commerce-specific filters, product matching, and search relevance.

Product recommendations

The Copilot can present suitable product suggestions when the configured flow and connected product or search services support recommendation scenarios.

Recommendations can be based on:

• The buyer’s stated requirements

• Product attributes

• Search results

• Customer-specific business rules

• Buyer or account context, where authorized

Recommendations are generated from the configured data and flow logic. They do not guarantee product suitability unless the relevant business rules and validation have been implemented.

Product comparison

The Copilot can help buyers compare products or product alternatives when the required product information is available.

A comparison can include:

• Product attributes

• Technical specifications

• Availability

• Prices

• Differences between selected products

• Customer-specific information, where authorized

The comparison is only as complete and accurate as the connected product data and the configured flow.

Product information

The Copilot can answer product-related questions using information from connected commerce services and approved knowledge sources.

Depending on the configuration, this can include:

• Product descriptions

• Technical specifications

• Product attributes

• Usage information

• Compatibility information

• Product documentation

• Related or alternative products

ICM and other connected source systems remain authoritative for the data they provide.

Price and availability information

The Copilot can present price and availability information when this data is available through the configured commerce services and the buyer is authorized to access it.

This can include:

• Product prices

• Customer-specific prices

• Availability information

• Applicable product or account context

The Copilot does not calculate or independently guarantee authoritative prices or availability. These values must be obtained from the connected commerce systems.

Basket operations

The Copilot can initiate supported basket operations when the required ICM APIs, authentication, flow logic, and PWA integration are configured.

Examples can include:

• Adding a product to the basket

• Adding a requested quantity

• Removing a product from the basket

• Clearing the basket

The exact supported operations depend on the configured flow and storefront integration. The resulting basket state remains managed by the commerce platform and storefront.

Order-template operations

The Copilot can support order-template scenarios when the relevant ICM capabilities and flow tools are available.

Examples caninclude:

• Creating an order template

• Adding products to an order template

• Adding quantities to an order template

• Retrieving information about an order template

These operations require the appropriate buyer permissions and customer-specific configuration.

Order and account information

The Copilot can retrieve buyer-specific order or account information when the buyer is authenticated and the configured integration forwards the buyer context to an authorized commerce service.

Examples can include:

• Order history

• Order-related information

• Account-specific information

• Customer or organizational context

The availability of this capability depends on authentication, token handling, ICM permissions, the configured flow, and the relevant API implementation. Buyer-specific information must remain isolated to the authorized buyer context.

Storefront navigation

The Copilot can request navigation to supported storefront destinations when the PWA integration supports this behavior.

Examples can include navigation to:

• The basket

• A product detail page

• An order or account area

• A service or information page

The exact destinations and navigation behavior are defined by the PWA integration.

Knowledge assistance

Knowledge assistance enables the Copilot to answer questions using approved, indexed customer content and other configured knowledge sources.

Approved customer documents

The Copilot can retrieve information from approved customer documents, including:

• Product documentation

• Service documentation

• Manuals

• FAQs

• Policies

• Procedures

• Other customer-provided documents

The documents must be prepared, indexed, and connected to the Copilot flow. The customer’s source system remains the authoritative source for the documents. The provided Azure Storage Account is an ingestion and indexing store, not a document-management system or source of truth.

https://intershop.atlassian.net/wiki/x/foA65Qs

Retrieval through Azure AI Search

Azure AI Search is the supported knowledge and vector-retrieval baseline for Copilot for Buyers.

Depending on the configuration, the retrieval solution can use:

• Text search

• Semantic search

• Vector search

• Indexed metadata

• Customer-specific filters

• Document or product-image embeddings

The retrieval result is provided to the configured IAB flow as context for generating the answer. Azure AI Search does not itself generate the final conversational response.

The implementation partner is responsible for customer-specific index configuration, ingestion, retrieval mapping, testing, monitoring, and retrieval quality.

https://intershop.atlassian.net/wiki/x/XoAs5Qs

Product and service documentation

The Copilot can use indexed product or service documentation to answer questions that are not fully covered by live commerce data.

Examples include:

• How a product is used

• Product or service procedures

• Technical explanations

• Frequently asked questions

• Customer-specific policies

• Supporting information related to a product

The flow should use retrieved content as the basis for the answer and handle missing, conflicting, or insufficient information conservatively.

Source references and provenance

The Copilot can display source references or provenance information when:

• Source metadata is retained during ingestion

• The retrieval tool returns that metadata

• The flow maps the metadata into the response

• The referenced source is accessible to the intended buyer

Source references identify supporting material. They do not by themselves guarantee that the generated interpretation is correct. Internal storage paths or administrative URLs must not be exposed to buyers.

Customer-specific knowledge sources

Customers and implementation partners can connect additional knowledge sources through configured retrieval tools.

Possible sources include:

• Customer-managed documents

• External retrieval services

• External vector stores

• Knowledge Graphs

• Other approved content systems

These are customer-specific extensions unless they are explicitly included in the supported Copilot solution. External vector stores and Knowledge Graphs are not interchangeable standard providers within the Copilot product. Their provisioning, integration, security, monitoring, and quality remain the responsibility of the customer or implementation partner unless separately agreed.

https://intershop.atlassian.net/wiki/x/foA65Qs

Image-related capabilities

Image-based product discovery

Where enabled, a buyer can submit an image to help identify or find a matching product.

The image-based flow can:

1. Receive an image through the Copilot interface

2. Route the image through the configured PWA integration

3. Invoke the image-recognition tool

4. Query the Intershop-managed image search API

5. Return matching product identifiers and related product information

6. Present the result in the Copilot conversation

The exact behavior depends on the image-routing implementation, the configured azureAiImageRecognition tool, and the target release.

Product-image indexing

Product images and associated metadata can be indexed for retrieval during image-based product discovery.

The current image-indexing process uses:

• Product images

• Product metadata

• Azure Storage

• An image-indexing service

• Image embeddings

• Azure AI Search

The indexing process can update the search index when new or changed images are available.

https://intershop.atlassian.net/wiki/x/HYAS5ws

Image similarity search

The image search service can compare an uploaded image with indexed product images and return similar or matching products.

Results can include:

• Product identifiers

• Associated product metadata

• Match information

• Confidence information, where provided by the service

The Copilot flow is responsible for interpreting the results and presenting them appropriately. Raw similarity scores must not automatically be exposed to buyers.

Image-based knowledge retrieval

Images can also form part of a broader knowledge-retrieval scenario. Depending on the indexing pipeline and configured retrieval tool, the Copilot can retrieve information associated with:

• Product images

• Diagrams

• Scanned documents

• Screenshots

• Other image-based content

This capability must not be described as generally available unless the relevant indexing, retrieval, PWA, and IAB components have been confirmed for the target release.

Storefront interaction

Copilot for Buyers can return both conversational content and structured results for the storefront.

Conversational answers

The Copilot can return natural-language responses based on:

• The buyer’s request

• The configured flow

• Connected commerce services

• Search results

• Retrieved knowledge

• Authorized buyer context

Product results

The Copilot can present product-related results when the flow connects to product data or commerce search.

The exact presentation depends on the Copilot Client and PWA integration.

Structured actions

The flow can return structured actions where the Copilot Client and PWA support the required action format.

Examples can include:

• Navigating to a storefront destination

• Adding a product to the basket

• Removing a product from the basket

• Opening a product or order context

• Initiating another supported storefront interaction

Structured actions must be limited to explicitly supported and validated PWA operations.

Navigation requests

The Copilot can request navigation to selected storefront destinations when navigation actions are configured.

The PWA and Copilot Client handle the resulting storefront behavior. The AI flow does not directly control the browser.

Basket and other PWA actions

Basket operations and other storefront actions require cooperation between:

• The IAB flow

• The relevant commerce API

• The Copilot Client

• The PWA integration

The exact action set is release- and configuration-dependent. A successful chat response alone does not prove that all storefront actions are available.

Conceptual architecture

Copilot Client and PWA Boundary

The Copilot client is deployed through the regular PWA development and release process. Flowise deployment does not deploy or validate PWA changes. Changes that affect both the conversational flow and the storefront user interface must be coordinated across both release processes.

Commerce Search

The Copilot can use the commerce platform's configured search provider, such as Solr or SPARQUE. Commerce search remains responsible for authoritative product filters and commerce-specific relevance signals. SPARQUE uses graph-supported search capabilities internally. This does not expose a general-purpose Knowledge Graph for configuration through the Copilot.

Knowledge Storage and Indexing

Intershop provisions a customer-specific Azure Storage Account and Azure AI Search service as the supported knowledge and vector-search baseline for Copilot for Buyers.

Ingestion Workflow

1. The implementation partner places source documents or prepared chunks in the provided Storage Account. Content can be uploaded manually using Microsoft Azure Storage Explorer or automatically through Azure Storage APIs.

2. The partner configures the required Azure AI Search data source, index, indexer, and optional skillset.

3. For a standard PDF workflow, an Azure AI Search indexer reads the files from Blob Storage, extracts indexable content and metadata, and writes the results to the configured index.

4. Built-in Azure AI Search capabilities can be used for common chunking and vectorization scenarios. Customer-specific preprocessing or complex chunking is implemented and operated by the partner.

5. The partner validates indexing results and configures the IAB retrieval tool. The Copilot for Buyers blueprint contains an example for querying the provided Azure AI Search service.

Source of Truth and Lifecycle

The provided Storage Account is an ingestion and indexing store, not the customer's source of truth. Authoritative documents, document governance, and source-document versioning remain in the customer's source system.

The implementation partner maintains the customer-specific knowledge base. This includes adding, replacing, and deleting content; defining retention; maintaining indexes and indexers; triggering or scheduling reindexing; and monitoring ingestion, capacity, and search quality. Intershop operates and backs up the provisioned infrastructure.

Capacity

Capacity requirements depend on the number and size of indexed chunks and images. Each indexed chunk counts as one document; splitting a source document into multiple chunks therefore increases the required index capacity. Contractual capacity allowances are defined in the applicable service description.

Retrieval Capabilities

Supported Standard

Azure AI Search is the vector-search component provided and supported by Intershop. The standard blueprint provides a straightforward vector-search starting point that partners can adapt to the customer's indexed content and retrieval requirements.

Partners can configure customer-specific indexes, indexers, skillsets, query behavior, and IAB retrieval tools. Intershop support covers the provided Azure AI Search infrastructure. The partner is responsible for the functional quality of the customer-specific retrieval solution.

Advanced Retrieval Pattern

Intershop also provides an optional advanced example that demonstrates a deterministic multi-stage product-search pattern:

1. Parse the buyer's request into search terms, filters, and intent signals.

2. Use text, semantic, and vector retrieval in Azure AI Search to identify product candidates and relevant descriptive context.

3. Apply authoritative commerce filters and relevance signals through the commerce-search service.

4. Fuse the retrieval signals, validate candidate suitability, and rerank the selected candidates with an LLM.

5. Apply configured quality and business rules before returning the final recommendations.

The available example uses Solr for the commerce-search step. An equivalent SPARQUE-based variant can be provided on demand for a customer project. The advanced pattern is an implementation example and is not enabled automatically in every Copilot deployment.

Alternative Vector Stores and Knowledge Graphs

A general provider abstraction for interchangeable vector stores is not part of the standard Copilot for Buyers product. Azure AI Search remains the mandatory component provided and supported by Intershop.

A partner may connect an external vector store through a customer-specific IAB tool. The partner is then responsible for provisioning, licensing, security, ingestion, lifecycle, integration, monitoring, compatibility, and retrieval quality. The external service is outside the standard Intershop support scope unless separately agreed.

Copilot for Buyers does not provide a native, managed Knowledge Graph. External Knowledge Graphs can be connected as customer-specific retrieval or context sources. The partner is responsible for the integration, permissions, query mapping, result transformation, and quality assurance.

Sources and Provenance

Copilot for Buyers supports visible source citations or provenance links in answers. Partners must preserve suitable source metadata during ingestion and map it to retrieval results. A displayed reference must point to a source the buyer is authorized to access and must not reveal internal storage locations or administrative URLs.

Access Control and Security Boundaries

Intershop grants the implementation partner only the Azure roles required for the agreed implementation and operational tasks. Unrestricted administrative access to the Azure subscription is not provided. Exact role assignments depend on the actions the partner must perform and follow the least-privilege principle.

Credentials and secrets must not be embedded in IAB code or exported flow definitions. They are maintained separately for each environment. Partners are responsible for protecting customer-specific credentials and for ensuring that connected source systems enforce the required authorization.

For further security requirements, see Concept - Security for Intershop Copilot for Buyers.

Environment and Release Model

Agent Builder Workspaces

Copilot for Buyers uses separate IAB workspaces for INT, UAT, and PRD. Development takes place in INT; testing and approval take place in UAT; and the approved configuration is then promoted to PRD. Production should not be edited directly.

Current IAB Promotion

The current process uses the standard IAB export and import functions for complete Chatflows and Agentflows. Flow-level nodes, prompts, and tools are transferred as part of the flow definition.

Credentials, secrets, and global or workspace variables are excluded and must be configured separately in every target workspace. After import, the partner must validate environment-specific dependencies, endpoints, permissions, and credentials before activating the flow.

Knowledge Infrastructure Environments

By default, Intershop provides one central Azure Storage and Azure AI Search setup for the customer solution. Separate INT and UAT knowledge infrastructures can be provisioned on demand. The existence of three IAB workspaces does not imply that three separate Storage Accounts and Azure AI Search services are provided automatically.

Azure AI Search configuration can be reproduced in another environment through Azure APIs or SDKs. Indexes, indexers, skillsets, data sources, and synonym maps are recreated with target-specific connections and permissions. Indexed content is rebuilt from the target source rather than treated as the authoritative deployment artifact.

Responsibilities

Limitations

• Azure AI Search is the supported vector-search baseline. Native multi-vector-store portability is not included.

• Customer-specific IAB tools, external vector stores, external Knowledge Graphs, and advanced retrieval adaptations are implemented and operated by the partner unless separately agreed.

• The ingestion Storage Account is not a document-management system or source of truth. Product-level authoring, approval, versioning, and records-management workflows are not provided.

• Retrieval and generated answers depend on source quality, source availability, permissions, configuration, and AI-model behavior.

• AI-generated answers and recommendations must be tested for the customer use case. Business-critical decisions should retain appropriate human review and escalation paths.

• The current IAB export/import does not clone credentials, variables, Azure knowledge infrastructure, or PWA configuration.