SAP data modernization

SAP to Microsoft Fabric Data Model Accelerator

Turn complex SAP data structures into analytics-ready Microsoft Fabric models.

Making SAP information useful across enterprise analytics takes more than extracting tables. Teams need to understand metadata, custom fields, business rules, relationships, incremental data movement, and trusted analytical models.

Our assessment-led approach combines SAP metadata discovery, source-aware extraction planning, schema mapping, transformation engineering, dimensional modeling, validation, and Power BI semantic model design. The right route depends on each SAP source, object, delta requirement, and licensing context.

Request a technical assessment

Automation scope, connector support, and delivery outcomes depend on the verified product capabilities and SAP landscape.

What Is the SAP to Fabric Data Model Accelerator?

It is an engineering-led modernization approach intended to support repeatable metadata analysis, source mapping, transformation preparation, and model construction for supported scenarios. A typical program discovers SAP systems and objects, selects appropriate extraction interfaces, maps technical structures and business meaning, creates curated data products, validates results, and establishes governance and operations.

Treat the accelerator as a way to support migration—not as a promise that every SAP object can be converted automatically. Custom ABAP, complex BW logic, source-specific extraction requirements, and business semantics require review by the appropriate technical and business owners.

SAP Systems and Data Sources We Can Assess

Each SAP landscape has different data models, extraction interfaces, customizations, and semantics. Validate source objects and integration options for the actual release and configuration.

SAP ECC

Assess standard and custom tables, append structures, ABAP CDS views, keys, relationships, change capture, extraction windows, and access controls.

SAP S/4HANA

Review the release, deployment model, CDS views, released interfaces, compatibility views, custom extensions, and delta requirements. Do not assume an ECC extraction design transfers unchanged.

SAP BW and BW/4HANA

Inventory ADSOs, CompositeProviders, queries, BEx variables, process chains, hierarchies, key figures, transformations, and existing curated outputs.

SAP HANA

Evaluate supported HANA objects and calculation views, including parameters, semantics, authorization, connector activity, gateway, and source impact.

ABAP CDS and custom extensions

Assess CDS exposure, custom Z-tables and fields, append structures, ABAP dependencies, and business-owner context before target mapping.

Choose a Supported SAP Connectivity Route

No single connector covers every SAP source and extraction requirement. Select a route based on source, object, delta behavior, Fabric workload, gateway, operational constraints, and licensing; verify current Microsoft and SAP support before implementation.

SAP Table connectors

Assess supported table and view extraction scenarios for the SAP release, required activity, gateway, authentication, delta behavior, and licensing.

SAP BW connectors

Evaluate BW Application Server, Message Server, or Open Hub paths. Connector availability can vary by Fabric activity; verify the current support matrix.

SAP HANA connector

Confirm the HANA version, supported objects and activities, client or driver prerequisites, network, gateway, authentication, and source impact.

OData and CDS

Review service exposure, authentication, pagination, filtering, rate limits, delta support, and whether the specific CDS view is released and accessible.

SAP CDC and ODP

Microsoft's guidance describes SAP CDC as an Azure Data Factory data-flow capability using ODP. Review SAP Note 3255746 and applicable licensing with the SAP team.

Partner and Datasphere routes

Where appropriate, evaluate a verified partner integration or existing Datasphere outbound pattern. Validate exact source, version, destination, and change-data support.

From SAP Sources to Fabric Data Products

A layered design can separate traceable source extracts from standardized and business-ready data. The Bronze, Silver, and Gold pattern should be tailored to the organization's governance, data contracts, and workloads.

Source and extraction
Choose the supported SAP interface and plan identity, delta, and source-impact controls.
Bronze: raw landing
Retain traceable source data with ingestion metadata and a documented reload strategy.
Silver: standardized
Normalize types, keys, relationships, quality rules, and reusable business logic.
Gold: analytics-ready
Publish curated domain data products for Fabric Warehouse, Lakehouse, and semantic models.

Power BI semantic models can sit over curated data. Direct Lake may suit eligible OneLake Delta tables, but it is not an SAP connector or a universal replacement for Import or DirectQuery.

What the Accelerator Helps Prepare

Potential support areas include metadata cataloging, schema and relationship analysis, candidate transformation scaffolding, model preparation, and reconciliation. The product's actual supported inputs and generated outputs should be confirmed against representative SAP assets.

Metadata discovery

Build a structured inventory of available tables, fields, data types, keys, relationships, custom objects, and dependencies.

Schema mapping

Propose target names and Fabric-compatible types, preserve precision and scale, and flag ambiguous or unsupported fields.

Relationship analysis

Identify candidate header/item relationships, business keys, fact and dimension structures, join paths, and modeling exceptions for review.

Transformation acceleration

Prepare candidate SQL, PySpark, and measure templates for supported logic. Custom ABAP and complex BW or HANA behavior need explicit assessment.

Fabric scaffolding

Prepare reusable pipeline and notebook patterns, layer conventions, schema documentation, validation queries, and deployment checklists.

Model and reconciliation support

Prepare semantic model definitions and checks for row counts, keys, financial totals, quantities, nulls, duplicates, and referential integrity.

Example SAP Data Model Patterns

These are illustrative domain patterns. Confirm actual SAP objects, keys, grain, relationships, and business semantics for the customer's release and configuration.

Finance and general ledger

Journal and balance facts with company, account, fiscal calendar, currency, and exchange-rate dimensions.

Validation: Reconcile debits and credits, posting periods, reversals, ledgers, and currency conversion.

Order to cash

Sales order, delivery, and billing data with customer, product, sales organization, and calendar dimensions.

Validation: Check quantities, delivery status, billing amounts, pricing conditions, cancellations, and units.

Procure to pay

Purchase orders, goods receipts, invoices, suppliers, materials, and purchasing organization.

Validation: Compare ordered and received quantities, open balances, invoice matching, periods, and currencies.

Inventory and supply chain

Inventory snapshots and material movements with material, plant, storage location, and movement type dimensions.

Validation: Validate stock balances, movement quantities, unit conversion, valuation, and snapshot timing.

Manufacturing and production

Production orders, operations, confirmations, work centers, materials, and plant-level data products.

Validation: Check order status, planned and actual quantities, scrap, yield, duration, and material consumption.

A Five-Stage SAP-to-Fabric Delivery Method

Move in manageable waves, starting with agreed source access, business priorities, and measurable validation criteria.

Stage 1
Discover and rationalize
Inventory SAP systems, source objects, custom extensions, volumes, data products, dependencies, and business owners.
Stage 2
Extract metadata and map
Capture available metadata, map fields and types, confirm keys and relationships, and document custom-object exceptions.
Stage 3
Design logic and models
Define target data products, transformations, semantic models, security, and suitable Fabric workloads.
Stage 4
Build and operate
Implement approved connectivity, ingestion, Lakehouse or Warehouse structures, data quality, monitoring, retries, and deployment practices.
Stage 5
Validate and transition
Reconcile source and target results, test access and recovery, benchmark workloads, obtain business acceptance, and plan cutover.

Preserve SAP Business Meaning

SAP data is more than a collection of tables. Configuration, application logic, organizational structures, and business conventions all affect its meaning.

Currency and exchange rates

Keep amount and currency context together; define rate type, conversion date, decimal precision, and rounding.

Units of measure

Retain quantity and unit fields together and define conversion rules before aggregating across materials or processes.

Fiscal calendars and time

Preserve fiscal variants, posting periods, special periods, time zones, and period-to-date behavior.

Keys and relationships

Confirm business keys, table relationships, header/item grain, and delete or late-arriving change behavior against the actual SAP configuration.

Security and custom logic

Review extraction authorizations, custom ABAP, Z-objects, and the target Fabric access model. SAP permissions do not transfer automatically.

Direct Lake suitability

Evaluate eligible OneLake Delta tables, data types, model features, security, capacity, and query needs. Direct Lake is not an extraction connector or universal default.

Common questions

SAP to Fabric Data Model Accelerator FAQs

What is the SAP to Fabric Data Model Accelerator?

It is an accelerator-led approach to assess SAP metadata, map source structures, prepare transformations, and build analytics-ready data models in Microsoft Fabric. The exact automated scope depends on supported SAP sources, object types, and verified product capabilities.

Which SAP systems can be assessed?

The assessment can cover SAP ECC, S/4HANA, BW, BW/4HANA, HANA, CDS views, and custom Z-objects. The supported extraction route and modeling approach depend on each source release, configuration, object, and business requirement.

Does Microsoft Fabric have one SAP connector for every workload?

No. SAP Table, BW, HANA, OData, CDC, partner, and Datasphere routes address different sources and workloads. Confirm current connector support, activity, prerequisites, gateway, authentication, and licensing for the specific scenario.

Does Fabric support SAP CDC natively?

Microsoft's SAP extraction guidance describes SAP CDC as an Azure Data Factory data-flow capability using ODP, with existing ADF pipelines able to integrate with OneLake. Do not assume every SAP CDC use case is available as a native Fabric Data Factory connector.

Are there SAP licensing considerations for ODP or CDC?

Potential restrictions may apply. Microsoft advises reviewing SAP Note 3255746; confirm permitted usage with SAP and the customer's licensing team before choosing an ODP-based extraction path.

Can the accelerator automatically convert ABAP or every SAP object?

Do not assume arbitrary ABAP programs, BW queries, or HANA calculation views are automatically converted. Supported constructs may be analyzed or used to prepare candidate code, but complex business logic requires engineering review and testing.

How are SAP currency and unit semantics preserved?

Keep amount and currency, or quantity and unit, fields associated; define conversion rates, dates, precision, rounding, and unit rules; and reconcile representative results with trusted SAP outputs.

Can SAP data be stored in OneLake as Delta tables?

Fabric Lakehouse data can use Delta tables in OneLake. Extraction and transformation still need to preserve required types, keys, business semantics, and data-quality expectations.

Can Power BI Direct Lake be used for SAP data?

Direct Lake can serve semantic models over eligible Delta tables in OneLake after SAP data is ingested and prepared. It has capacity, data-type, security, and feature considerations; evaluate it against Import, DirectQuery, and composite options.

How is SAP-to-Fabric accuracy validated?

Compare source and target row and distinct-key counts, financial totals, quantities, referential integrity, representative records, and incremental changes. Account for timing, rounding, currencies, units, and business rules, then obtain business-owner sign-off.

Will moving SAP data to Fabric always lower costs or improve performance?

No. Outcomes depend on extraction, SAP source impact, capacity, data volumes, modeling, refresh, query patterns, licensing, and parallel-running costs. Establish a baseline and test representative workloads.