Skip to main content

Ingest and reconcile

Onboard a new supplier in days rather than weeks

Fuse reads the file or the API, infers its structure, proposes the mapping to your model and flags blocking errors before they reach the ERP, the PIM or the shop.

Request a demo

The situation today

What goes wrong today

Every supplier sends a different shape. Excel from one, CSV from another, XML, a REST endpoint or a FAB-DIS or BMEcat catalogue from the next. Internal teams are no more consistent: each one writes product information to its own logic, so formats, standards and conditions drift. Then each application holds its own model, because the ERP, the PIM and the shop were each built for a different job. All of those models have to be reconciled before any of the data is usable, and today that reconciliation happens in a spreadsheet, by hand, every time a file arrives.

The mechanism

How it runs in Fuse

The supplier drops a file, exposes an endpoint or sends a catalogue in an exchange standard such as FAB-DIS, ETIM xChange, BMEcat or GS1. Fuse identifies the structure and the columns, then proposes a mapping to your internal model, which you correct where it guessed wrong. Units, formats and values are normalised on the way through. Blocking errors are reported before any row moves on, with the source row and the rule that failed. Validated records are then sent to the PIM, the ERP or the shop.

  1. 1

    Collect

  2. 2

    Understand

  3. 3

    Reconcile

  4. 4

    Enrich

  5. 5

    Control

  6. 6

    Publish

Illustrative payloads. Field names match the default Fuse model.

Inputs are files (CSV, TSV, Excel, JSON, XML), REST, connectors and exchange standards. A PDF catalogue or a scanned price list is not an input, so a supplier who only sends PDFs stays a manual step.

What changes

What you get

  1. Integration measured in days

    A new source becomes a mapping exercise rather than a project. The structure is proposed; your team corrects and confirms it.

  2. No second Excel pass

    Normalisation runs inside the pipeline, so nobody re-cleans the same file after every drop.

  3. Collections go live sooner

    The bottleneck moves from data preparation back to the commercial decision about when to publish.

  4. One quality standard for every source

    The same rules apply to a supplier file, an internal extract and an API. Quality stops depending on who sent the data.

Indicators to track

How to measure it

Time to integrate a new source

days from first file to validated records

Rows imported without intervention

percentage of rows per import

The clearest signal that the mapping and the rules are right.

Errors detected before integration

blocking errors per import

Time from file received to data available downstream

hours to the PIM, ERP or shop

Next

Related use cases

  • Duplicate detection

    Fuse compares identifiers and mapped fields, surfaces the records that describe the same product, explains the proposed match and keeps the source trail.

    Ingest and reconcile

  • Attribute normalisation

    Fuse converts source values into the reference lists your business uses for units, colours, sizes, materials, dimensions and technical characteristics.

    Speak one product language

  • Product golden record

    Fuse reconciles ERP, PIM, supplier and team data, applies the priorities you set, and produces a traceable record every system can read.

    Ingest and reconcile

Send us the supplier file that costs you the most time

We map it, run it through Fuse and show you the structure it proposes and the blocking errors it stops before they reach your systems.

Request a demo

Thirty minutes on your file, not a slide deck.