Expertise · process automation

Product data automation: the basics

Recurring data validation, transformations and deliveries can often be supported by defined rules. The key is knowing which steps are unambiguous and where people need to review or decide. We help identify suitable workflows and their limits.

Explore automation possibilities
Schematic illustrationRecurring process
TriggerSource data
Mapping and rules
ProcessingTransformationValidation
Review and approvalOpen issuesApproval

Recurring data validation, transformations and deliveries can often be supported by defined rules. The key is knowing which steps are unambiguous and where people need to review or decide. We help identify suitable workflows and their limits.

Good candidates for automation

Source data

  • recurring imports and exports

Processing

  • validation of catalogues and supplier data
  • mapping source fields to target fields
  • normalisation of units and values

Output

  • generation of BMEcat, CSV, XML or platform output

Review and approval

  • approval and review processes

Exceptions and questions

  • monitoring, status messages and error handling

How we define the process scope

Fictional example

Example process: prepare lengths for output in millimetres

The input determines the next step

Trigger · New data version available

Are value and unit evidenced and covered by the agreed rule?

Yes · Normal case

Evidenced input: 2.4 m

Convert the unit using the agreed rule. The result can proceed to the defined next validation step.

No · Exception

Input: 2.4 · Unit missing

Do not guess a target unit. Conversion of this value remains suspended; refer the question to the data owner.

Separate path · Expert clarification

Data owner clarifies the source → provides evidenced input → checks it against the rule again. Without evidence, the exception remains unresolved. No automatic continuation is promised.

The example explains the boundary between rule-based processing and an expert decision. It promises neither full automation nor ongoing operation or support.

Within the agreed approach

  1. define the trigger, input data and target result
  2. make manual decisions and exceptions visible
  3. identify systems, formats and interfaces
  4. distinguish steps that can be automated from those needing expert review
  5. define acceptance, monitoring and responsibility

Managed service or custom solution?

A managed service fits when excodata takes over a clearly defined workflow on a recurring basis. A custom solution fits when systems and rules need a customer-specific technical connection. Development, operations and further development are not automatically bundled together.

Limits

Agreed limits

  • no general digitalisation consulting without a specific process
  • no full automation where expert review remains necessary
  • no ongoing custom development without a defined scope
  • no guarantee that every exception can be automated

FAQ

Does the process need to be fully documented already?

No. A concrete starting point, the systems involved, typical inputs and the desired result are enough for an assessment. The scope is refined before the offer.

Can a small manual workflow be automated?

Yes, if the recurrence, effort spent on errors or time required justifies implementation and the process is clear enough.

Who operates the solution after implementation?

This is agreed separately. A development project does not automatically include ongoing operations or support.

Expertise · process automation

Define your task

Explore automation possibilities