Custom solution · automation

Automate your data process

You want a technical implementation of a specific manual data workflow. We connect the required systems and rules into verifiable automation. Specialist decisions, exceptions and errors remain traceable.

Schematic illustrationFrom workflow to solution
TriggerSource dataCustomer / source system
Mapping and rules
Processingexcodata · agreed scopeTransformationValidation
Review and approvalOpen issuesApproval

Recipient / target system · Acceptance

You want a technical implementation of a specific manual data workflow. We connect the required systems and rules into verifiable automation. Specialist decisions, exceptions and errors remain traceable.

Suitable starting points

Source data

  • files are regularly imported or exported manually

Mapping and rules

  • data needs validating and transforming according to defined rules
  • several systems are connected through intermediate Excel steps

Review and approval

  • approvals and exceptions run through unclear email processes

Output

  • catalogues or platform data are generated repeatedly

Exceptions and questions

  • errors are detected late and are hard to trace

From workflow to solution

Fictional example

Fictional acceptance cases · Rule R-4, version 1 · Length to millimetres

One rule, two inputs, two verifiable results
R-4: For an evidenced numeric value with unit “m”, multiply by 1000. If the unit is missing, produce no target value and flag the record for expert clarification.
Test case and input
T-01
Value
2.4
Unit
m
Expected target value
2400 mm

2.4 × 1000 = 2400

Verifiable behaviour
T-01: target value 2400, target unit mm; no clarification for a missing unit.
Test case and input
T-02
Value
2.4
Unit
Missing
Expected target value
No target value
Verifiable behaviour
T-02: no target value; log “unit missing” with identifier T-02 for clarification. No replacement value 0 and no assumed unit.
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.

Only two fictional cases for the rule branch shown. Other units, invalid numbers, interfaces, security requirements and retries need separately agreed cases. Passing one test is not overall approval or a recipient acceptance guarantee.

Within the agreed approach

  1. define the trigger, input data and goal
  2. identify rules, decisions and exceptions
  3. clarify interfaces and security requirements
  4. define required expert review and the limits of automation
  5. implement processing, logging and error paths
  6. accept the result using realistic test cases

Technical possibilities

Depending on the agreed scope, we connect interfaces, file processing, background services and scheduled workflows. Where people need to review or approve, that step is built into the process.

Limits

Agreed limits

  • no full automation where rules are missing or specialist uncertainty remains
  • no general digitalisation consulting without a specific target process
  • no ongoing operations without a separate agreement
  • no unlimited extensions beyond the accepted scope

FAQ

Can a process be automated gradually?

Yes. A narrow, measurable subprocess is often the better start when interfaces or exceptions are not yet fully clarified.

Can expert review remain part of the process?

Yes. Review points can be a deliberate part of the solution where decisions cannot be reliably automated.

Which technology does excodata use?

The choice depends on the system environment, operations and scope. What matters is a reliable workflow and testable results.

Custom solution · automation

Define your task