Know which SAP flows need attention. Bring failures, performance, and the next step into one clearer view.
Start with a test tenant. You choose the flows and when monitoring starts.
A QUIETER WAY TO KEEP WATCH
Try a filter, then select a flow to explore its details.
Most SAP integration work isn't building iFlows.
It's understanding them. Monitoring them.
Finding failures before the business does.
CPI Copilot means less time searching dashboards, and more time solving the actual problem.
Steps helps you navigate a complex iFlow and locate a palette on the canvas. Trace brings captured SAP errors and data observations together with practical next checks.
Confirm whether the receiver requires this field, then compare its exact path with earlier captures. An empty field alone does not prove a failure.
A red row in the log is the easy case. These are the ones that reach the business first.
A scheduled iFlow never ran. Nothing technically failed, so nothing alerted anyone — and your FAILED count still says zero. You find out when the business calls.
Your integrations still succeed. But today's runtime is 3x yesterday's, messages are piling up in retry, and nothing in the dashboard is red yet.
Open the MPL. Open the log. Open the package. Open the artifact. Cross-reference. Repeat — every single time something breaks.
Built around the work you actually do — understanding and operating your integrations.
Read a large iFlow in true execution order and jump directly to the step you need.
Review available captures from one message and investigate its errors and data changes.
See what changed between two iFlow states without reading raw BPMN by hand.
Plain-English answers from live data.
Watch only the integrations you care about — and hear about it the moment they break, by email.
See what's stuck in your queues — and know exactly what to do about it.
Upload a ZIP — no tenant required.
Ask for any report and have it emailed.
Know what got slower, and which step did it.
Draft an iFlow structure and supporting scripts, then review and test them against your requirements.
We're still refining accuracy. It's strongest on standard patterns — complex custom mappings still belong to you.
Three steps, one read-only credential, nothing installed in your tenant.
A monitoring-only OAuth client in your tenant. Least access possible.
Save it in Settings and check that SAP access works before enabling monitoring.
Choose the integrations that matter, review alert recipients, then press Start monitoring.
Review the data paths and test with a non-production tenant first.
It reads message logs and metadata. There is no code path that lets monitoring change anything in your tenant.
Build and management actions can change SAP artifacts when you request them. Review the target and proposed change before proceeding. Monitoring and the browser extension do not edit artifacts.
Saved tenant credentials are encrypted. The backend decrypts them when it needs to authenticate your requested SAP operations; secret values are not returned in settings responses.
Cloud monitoring reads errors and operational metadata. Trace payloads are inspected inside your SAP browser tab and are not sent to our backend or an AI provider.
Prefer your own Azure OpenAI, Anthropic or Gemini key? Use it. Your key, your bill, your data path.
Monitor the flows that matter first. Check duration depends on SAP response times and data volume. Validate performance on your own tenant before expanding coverage.
CPI Copilot sits on top of SAP Integration Suite — it doesn't replace anything.
| Today | With CPI Copilot |
|---|---|
| Search through dashboards | Ask in plain English |
| Read the MPL manually | Failure summaries and optional AI-assisted analysis |
| Drag around a large iFlow canvas | Execution-order list with click-to-locate |
| Compare trace steps by hand | Captured findings and practical next checks |
| Build reports by hand | Reports generated and emailed |
| Hunt for which package it's in | Instant discovery |
| A missed run nobody notices | An alert before the business calls |
| A queue quietly backing up | Diagnosed — retry or move, in one click |
| Raw runtime numbers | What changed, and which step did it |
The public beta is free. Flat pricing comes later — your message volume will never change what you pay.
No credit card required
Enterprise plans coming soon
Connect a test tenant, choose a few critical flows, and review your alert settings before starting monitoring.
Questions about setup, facing an issue, or have feedback? Email us.