Microsoft Dynamics 365 Integration - Signalx Webflow Ecommerce Website Template

Dynamics 365 isn't one product — it's a family. D365 Finance & Operations (formerly AX) is for mid-market and enterprise manufacturers; D365 Business Central (formerly NAV) is for smaller operations. Different architectures, different integration approaches, both supported with their own connector.

D365 F&O vs D365 Business Central — different products

Buyers sometimes assume "Dynamics 365" is one product. It isn't. D365 Finance & Operations (D365 F&O) is the descendant of Microsoft Dynamics AX — a full-featured enterprise ERP serving mid-market and enterprise manufacturers. D365 Business Central is the descendant of Microsoft Dynamics NAV — a smaller-footprint ERP serving smaller manufacturers and divisions.

They share branding and some surface concepts but are architecturally distinct, with different data models and different integration mechanisms. Horizon has separate connectors for each.

D365 Finance & Operations integration

Connection approach

F&O integration uses three complementary mechanisms:

Dataverse (Common Data Service) — for entities exposed through Dataverse and Power Platform. Microsoft's preferred modern integration approach.
OData APIs — for direct access to F&O entities (Items, On-hand Inventory, Sales Orders, Purchase Orders) where Dataverse access isn't configured.
Data Management Framework — for bulk historical data and entities not exposed through Dataverse or OData. Standard import/export framework, scheduled exports to Azure Blob Storage that Horizon picks up.

F&O-specific data objects Horizon reads

Released products with their planning groups, dimension groups, and storage dimensions
InventDim records — the dimensional inventory model that's distinctive to F&O
On-hand inventory by site, warehouse, location, batch, and serial
Bills of material (BOM and BOMVersion tables)
Routings (Route, RouteOpr) with operations and resources
Resources, resource groups, and capacity calendars
Sales order header and line history
Purchase orders and confirmed receipts
Vendor master with lead times

F&O-specific gotchas the connector handles

Dimension groups and inventory dimensions — F&O's flexibility in defining what makes an item "different" (size, color, batch, location, etc.) is unusual among ERPs. The connector reads dimension groups explicitly rather than assuming a standard set.
Site vs warehouse vs location — F&O has three nested location concepts where most ERPs have one or two. Horizon maps these explicitly during setup.
Production journals — for customers using F&O's manufacturing module, production journals are how completed work orders feed back into inventory. The connector captures these for accurate inventory consumption.
Item allocation keys — F&O's allocation key concept is used for forecasting and aggregation. The connector respects existing keys rather than imposing Horizon's planning hierarchy on top.

D365 Business Central integration

Connection approach

BC integration is REST-based using the standard BC API:

BC REST API — primary integration path. Items, Item Variants, Inventory, Sales Orders, Purchase Orders all accessible through standard endpoints.
OData v4 — for BC entities not in the REST API, OData provides broader coverage.
Custom API extensions — for customer-specific data needs, BC's extensibility model allows custom APIs that Horizon can call.

BC-specific data objects Horizon reads

Item records with stockkeeping units (item-location combinations)
Item Variants for size/color/configuration variations
Locations and Bins for inventory placement
Production BOMs and Routings (for customers using the manufacturing module)
Sales lines and sales shipment history
Purchase lines and receipts

BC-specific considerations

Stockkeeping Units (SKUs) — BC's SKU concept is item-location, similar to NetSuite. The connector treats these as the planning units rather than just item-level.
Manufacturing module is optional — Many BC customers don't have the manufacturing module enabled. The connector adapts to what's actually licensed and configured.
Multi-company setups — BC's multi-company structure works differently from F&O's multi-legal-entity setup. The connector handles BC's per-company database approach explicitly.

Write-back to D365 (both versions)

Planned purchase orders
Planned production orders (F&O) or production journal entries (BC)
Item planning parameter updates — reorder point, safety stock, minimum/maximum inventory levels
Demand forecast — written to F&O's demand forecast tables or BC's planning worksheet, depending on version

Microsoft Fabric as the integration layer

An increasing number of D365 customers are adopting Microsoft Fabric as their unified analytics and data layer. For these customers, Horizon connects directly to Fabric's OneLake rather than to D365 itself. This pattern is faster because the data is already aggregated and the customer's data team has already solved the permissions and schema problems.

Horizon has production deployments using Microsoft Fabric as the integration layer for both D365 F&O and D365 BC environments. The Fabric path typically reduces integration time by 2-3 weeks compared to direct D365 connection, when Fabric is already populated.

Alternative integration path

For environments where direct API or Dataverse access isn't practical, D365's Data Management Framework can produce scheduled exports to Azure Blob Storage or SFTP that Horizon ingests. Used alongside the API paths for hybrid deployments.

Authentication

D365 F&O and BC: OAuth 2.0 with Azure AD service principal. The customer's Microsoft 365 administrator creates the app registration and grants the appropriate permissions.
Microsoft Fabric: Service principal with Fabric workspace access. Standard Microsoft authentication pattern.
Data Management Framework exports: Service principal with access to the Azure Blob Storage container where DMF exports land.

Representative reference deployments

An anonymized $750M discrete manufacturer running D365 F&O across 5 plants deployed Horizon for demand planning, inventory optimization, and finite capacity scheduling. The integration used Dataverse for most entities and Data Management Framework for bulk historical demand. Project ran 11 weeks end-to-end; the F&O integration completed in week 7. The customer's IT team spent roughly 60 hours on the integration — most of it on Azure AD app registration and Dataverse entity exposure.

A separate anonymized customer — $200M consumer products manufacturer on D365 Business Central — deployed Horizon for demand planning in 6 weeks. BC's REST API made the integration straightforward; the bottleneck was master data quality (item categorization that hadn't been maintained), not the connector.

Power Platform integration

For customers heavily invested in Power Platform, Horizon's planned orders, forecasts, and exception flags can feed Power BI dashboards directly. Power Automate flows can trigger Horizon actions (re-forecast, re-optimize) based on D365 events. This integration is configured per customer rather than out-of-box.