Skip to content
digitapulia.com
Continuity and compliance

Integration monitoring

We find out that a data flow has stopped when a customer complains.
What we do
We set up automatic checks that watch the data flows between your systems continuously and send alerts when something stops, and we give a first response within the hours set in the contract.
How we do it
First we list with you the flows that matter: orders coming from the online store or a marketplace into the ERP, invoices going out, shipments passed to the carrier. For each one we write down what working means: how often the data should move and which errors can be seen. Then we set up the checks, which look at each flow continuously, and the alerts, which reach the people you choose and us when a flow stops or piles up errors. Within the hours written in the contract we give the first response: we see what stopped and why, start it again when doing so puts no data at risk, or tell whoever runs that system what they need to do. Every month we send you a report on what happened.
What you receive
A status panel for your flows, and a monthly report.
What changes for you
When one of the monitored flows stops, you hear about it from an alert and not from a customer complaint. Problems that keep coming back show up in each month's report, with their cause, instead of being fixed each time as if they were new.
What we need from you
The list of your systems and of how they connect, access to read the status and errors of each flow, the permissions for the first response agreed with you, and the names of the people who receive the alerts.
Who it is for
IT and operations
How you buy it
For a recurring fee: the fee is written down before we start and we give it to you when you ask.

Example use cases

These are examples: they do not describe a real client.

  • A cosmetics maker in Italy that sells on a marketplace in Spain and Mexico, with orders connected to its ERP

    The situation

    Marketplace orders enter the ERP through a connection, and from there go out to the warehouse. One weekend the connection stopped bringing in orders because the credential it connected with had expired. No one saw it: the marketplace kept selling, the warehouse did not ship, and they found out from customers' complaints and from a marketplace notice about late shipments.

    What we do

    With the online sales team we write down what it means for the connection to work: marketplace orders appear in the ERP shortly after each sale. The check compares the two sides and, if orders stop coming in, the alert reaches the person in charge of the online store and us. Within the agreed hours we find the cause, renew the credential with whoever manages it and check that the pending orders come in. The monthly report records it.

  • An industrial spare parts distributor in Spain with customers in Colombia and Ecuador, which receives orders by EDI and sends e-invoices

    The situation

    Large customers send their orders by EDI, and the distributor issues its e-invoices from the ERP. When a customer's system rejects an invoice, for instance because the order number is missing, the notice stays in an inbox no one checks, and the customer does not pay because, for them, the invoice never arrived. The accounts team finds out weeks later, when reviewing outstanding payments.

    What we do

    We put a check on each flow: incoming EDI orders, outgoing invoices and rejection responses. When an invoice is rejected or an EDI order cannot be read, the alert reaches the accounts team and us. Within the agreed hours we read the reason for the rejection and tell the accounts team which detail to correct. The status panel shows which flows are working and which are not, as of the latest check.

Frequently asked questions

Is someone on watch at all hours?

The alerts go out at any hour. The first response is given within the hours written in the contract: we do not promise that someone will be on watch at all hours.

Which flows can be monitored?

The ones we agree with you at the start, after seeing what can be read from each: if a system shows neither its status nor its errors, we tell you before including it.

What does the monthly report say?

Which flows stopped, when and for how long, why, what was done, and which problems keep coming back.