Getting started with ml-connector

Step 1: Create your workspace

Sign up and you get a free sandbox immediately. No card, no trial countdown. The sandbox is there so you can point us at your actual systems and find out whether we fit before any money changes hands, which is a far better test than a demo against sample data.

Step 2: Connect a system

Add the credentials for a system you want to link. They are encrypted the moment you save them and are never shown again. Two things make this step less painful than it usually is: every connector ships a setup guide telling you exactly where to find its credentials in that vendor's admin screens, and a verify button runs a live call against the system so you find out immediately whether the credentials work, rather than at 3am when a sync fails.

Step 3: Build a flow

A flow is one job: what to read, what to write, and how the fields line up between the two. Pick the systems, map the fields, and set a schedule. Complete field mappings are part of what we build for you, so this is not an exercise in reverse engineering someone else's field names.

Step 4: Watch the first run

Run it once by hand before you trust the schedule. You get the run history, the records it touched, and the audit entries it wrote. Watching the first run end to end is the fastest way to catch a mapping that is subtly wrong while it is still cheap to fix.

When something fails

Integrations fail, usually because a system on the far end was down or a record was not what it claimed to be. The product is built for that rather than surprised by it:

• Failed jobs retry automatically, with harder limits on high-stakes flows like bills and payments
• Records that still cannot be processed land on an errors screen instead of vanishing
• You can replay a failed record once the underlying problem is fixed
• Retries are deduplicated, so a replay cannot double-post an invoice or an order
• Alerts can be sent by email when something needs a human

Going live

When you move to a paid plan we provision a dedicated environment for you: your own database, queue, worker, and encryption key. Your flows move onto a real schedule with monitoring behind them. Nothing about the flows you built in the sandbox has to be redone.

If your system is not on the list

Our catalogue runs to over 1,700 systems, and 120 of them have a finished connector today. If yours is not one of them, that is a build, not a refusal: we write the connector into your own environment. Start a chat and describe what you need to move.

Common questions

Do I need a credit card to start?

No. The free sandbox needs no card, and you can point it at your real systems to test whether we fit.

How long does a first integration take to set up?

Connecting a system and running a first flow by hand is usually an afternoon. The slow part is normally getting credentials out of the vendor's admin screens, which is why every connector ships a guide for that.

Do I have to build the field mappings myself?

No. Complete field mappings are part of what we build for you. You can review and adjust them, but you are not starting from a blank grid.

What happens to my sandbox work when I upgrade?

It carries over. Upgrading provisions your dedicated environment and puts your flows on a real schedule. You do not rebuild anything.

Can I test against production systems safely?

Yes, and we recommend it. Run flows manually first and watch the run history before putting anything on a schedule.