Silent failures
An order that never reaches the ERP makes no noise. Nobody misses it until the customer asks.
Module 06 · Integrafy Platform · Operational control
ObserveIf you cannot see it, it is not automated.
A real-time dashboard of every sync: events, errors, latency and the status of each connector. Failures are detected before they reach the customer.
Events processed today48,312+6%
Error rate0.4%stable
Average latency1.8 sp99 4.2 s
Active connectors6/6all online
Connector status
SAP Business OneOnline0.9 s
PrestaShop B2BOnline1.2 s
Odoo · warehouseOnline2.4 s
Supplier FTP 12Retrying—
ShopifyOnline1.0 s
HubSpotOnline1.6 s
Events per hour
The challenge
When nobody watches the integration, the customer is the alert.
Automation frees the team from repetitive tasks. But if nobody sees what happens inside, a small failure can last for days.
The connector down over the weekend
The supplier changes their FTP on a Friday. On Monday there are three days of catalogue not updated.
Growing queues
A slow system piles up pending events and nobody knows how many there are.
What happened to this order?
Answering means opening three systems and asking two people.
Observe is built to answer in one click what came in, what went out, what failed and who was told.
What really matters
An integration is not judged when it works. It is judged when it fails.
That is why Observe is not a month-end report. It is the tool that tells you, in the moment, what failed, where and what to do.
Integrafy Observe
Operational control of the whole platform.
Observe logs every event that passes through Integrafy Platform: Connect syncs, Flow executions, changes in Data and agent actions.
It is the band that runs across the whole diagram: not one more module, but the way to see all the others.
In one sentenceIf you cannot see it, it is not automated.
Where Observe sits in Integrafy Platform
Included in every product. Higher plans add advanced dashboards, longer history and an SLA.
Real-time dashboard
The full state of your integrations, at a glance.
- Events processed, error rate, average latency and active connectors, updated in real time.
- Status of each connector: online or down, last event, response time and recent errors.
- From the dashboard to the failure in one click.
Events processed today48,312+6%
Error rate0.4%stable
Average latency1.8 sp99 4.2 s
Active connectors6/6all online
Connector status
SAP Business OneOnline0.9 s
PrestaShop B2BOnline1.2 s
Odoo · warehouseOnline2.4 s
Supplier FTP 12Retrying—
ShopifyOnline1.0 s
HubSpotOnline1.6 s
Events per hour
Event log
What happened to this order? The answer, in one search.
Every sync is logged with its origin, destination, duration and, if it failed, the reason. Search by order, reference or customer.
| Time | Object | Origin → destination | Duration | Status | Reason |
|---|---|---|---|---|---|
| 10:42:07 | Order 4500012 | PrestaShop → SAP B1 | 1.1 s | OK | |
| 10:41:55 | Stock A-4471 | SAP B1 → PrestaShop | 0.8 s | OK | |
| 10:41:31 | Supplier 12 catalogue | FTP → Integrafy Data | — | Retry 2/5 | Connection timeout with the FTP |
| 10:40:12 | Customer C-2031 | PrestaShop → SAP B1 | 1.4 s | OK | |
| 10:39:48 | Order 4500009 | Shopify → Odoo | — | Rejected | Customer without available credit · pre-validation |
| 10:39:02 | Price list T-2026-N | SAP B1 → PrestaShop | 2.1 s | OK |
Configurable alerts
Alerts when action is needed. And only then.
Latency thresholds, connector disconnection or retry-queue size. Every alert arrives with the link to the event that triggered it.
If
Latencyrises above the threshold
Connectordisconnects
Retry queueexceeds the set size
Errorsrepeat in a flow
Observe
- Evaluates the threshold
- Groups repeated alerts
- Attaches the link to the event
Notifies via
Emailto the owner
Slack · Teamsvia webhook
Pushon mobile
Step by step
What happens when a connector fails on a Saturday.
ERP stops responding
Connect retries
Observe watches
Owner person
1 Tries to send an order: the ERP does not respond
2 The event goes to the retry queue, with backoff
3 The queue grows above the set threshold
4 Alert via Teams or email, with the link to the event
5 The ERP is available again
6 The queue is processed with no duplicates (idempotency)
7 Closes the alert: what happened, when and how long it lasted
No order lost or duplicated · on Monday, the log explains the whole weekend
Historical metrics
Twelve-month trends. And your data, in your BI.
12 months of history by default, extendable: events, latency evolution and error distribution.
CSV export and REST API for Power BI, Tableau, Looker or Grafana.
Useful to prove what has improved, not just to put out fires.
Events processedAverage latency
Comparison
Checking logs when something fails, or knowing beforehand.
Almost every integration has logs. The difference is whether anyone looks at them and whether they answer business questions.
Who sees it Scattered technical logsOnly whoever knows where they are and how to read them. Integrafy ObserveOperations, eCommerce and management, each with their own view.
When it is detected Scattered technical logsWhen the customer complains. Integrafy ObserveWhen a threshold is crossed, with an immediate alert.
Search Scattered technical logsBy date and file, in each system. Integrafy ObserveBy order, reference or customer, in one place.
Context Scattered technical logsOne technical error line. Integrafy ObserveOrigin, destination, duration and reason, with the full event.
History Scattered technical logsWhatever has not been deleted. Integrafy Observe12 months by default, extendable.
Business Scattered technical logsUseless in a meeting. Integrafy ObserveExportable to your BI and comparable month to month.
How it is contracted
Observe is not bought separately. It comes with every product.
Everything that runs on Integrafy Platform is visible in Observe from day one.
Monitoring dashboard
Events, errors, latency and connector status, with email alerts.
IncludedAdvanced Observe
Advanced dashboards, extended history and export to your BI.
Higher plansContractual SLA
Guaranteed response times for incidents.
EnterprisePrices exclude VAT. Full details in the Integrafy Platform price list.
See the full price list →Frequently asked questions about Integrafy Observe
Do I need technical access to use Observe?
No. The dashboard is designed for non-technical profiles: operations management, eCommerce managers, warehouse leads. Everything is visual, with filters, search by order or reference and configurable alerts.
Does Observe keep a history of events?
Yes. Every event processed by Integrafy is logged with time, origin, destination, duration, status and, if it failed, the reason. History is kept for 12 months by default, extendable on higher plans.
Can I export Observe data to my BI?
Yes. Observe exposes a metrics REST API and CSV export for Power BI, Tableau, Looker or Grafana. Useful to compare month to month and prove what has improved.
Can alerts be sent by email or Slack?
Yes. By email, by webhook to Slack or Teams, or by push notification. They are configured by latency threshold, connector disconnection, retry-queue size or repeated errors, and every alert arrives with the link to the event that triggered it.
What metrics does the dashboard show?
Events processed, error rate, average and p99 latency, active connectors, the status of each connector with its last event and response time, retry queue and events per hour. All in real time.
Is Observe contracted separately?
No. The monitoring dashboard with email alerts is included in every Integrafy Platform product. Higher plans add advanced dashboards, longer history and BI export; the contractual SLA is Enterprise.
Demo on your case
We show you your integrations like you have never seen them.
A 30-minute session: which systems you have connected today, what is getting lost on the way and how it would look in Observe.
Request a demo