Your client shouldn't be the one who tells you it broke.

{{BRAND}} is run by {{FIRST_NAME}} in Dhaka, Bangladesh. We're new, so we have no customer list to show you. The free check is how you decide: you see the work on your own clients' workflows before you pay anything. We work Monday to Friday, during the Dhaka afternoon, so the daily note is ready by about 9am UK time, before the US working day starts. In Australia it lands in your early evening.

As little as possible, and you choose how. Either add one named person from our side with a view-only or monitoring role, where your plan has one, or forward the failure emails Zapier, Make and n8n already send to a desk address. The email route catches failures but not quiet or switched-off workflows, so we'd also ask for a short weekly export of run counts. On some plans an extra seat costs money, and some only offer full access; in either case the email route costs nothing and we'd rather you use it. We never need your API keys, your clients' app logins, passwords or billing details, and you can remove our seat any time.

Run histories can hold your clients' customer data. We read status, run times, step names, error messages and counts, and we don't need the contents of each record. Where your platform lets you, switch off saving of run data for those workflows or mask it. What we keep is held in our Google Workspace account with 2-step login, seen only by the person working on your account, never shared or sold, and deleted within 7 days if you leave. If personal data does reach us, we handle it as your processor (or sub-processor, where you work for your client) and will sign a data processing agreement. Your contract with your client may require their OK before you share run logs with us. We're happy to sign your NDA first. Workflows that handle health information are out of scope: please leave them off the list. Ask for our one-page data promise.

People, with good tools. Software helps us read long run histories, sort errors into types and draft the report. A person checks every flagged item against what you told us the workflow should do, and reads every note and report before it reaches you.

Each working day, for every workflow on your list: runs that failed or stopped, runs that happened less often than usual or not at all, workflows that are off but should be live, connections that need signing in again, and plan usage close to its limit where your seat shows it. For up to 3 workflows per client, we also compare a count you can export, such as rows in a form sheet against new records in the CRM for the same day. We can only check what the platform and your exports show us. Green runs with a wrong result are caught only where there's a count to compare.

No to both. You decide every fix, and your clients never hear from us unless you ask. We replay a failed run only on workflows you've marked safe, once, and we tell you the result. A replay can send a second email or text, or create a duplicate record, so the default is to flag it and let you decide. We never edit a workflow, switch one on or off, reconnect an app or change a setting.

Much of it, yes. The platforms' failure emails, free watchdogs and paid tools for agencies will alert you when a run fails, and some flag quiet workflows or produce a client report. What they don't do is have a person read the history each day against what each workflow is meant to do, check the counts you care about, and write a report you'd forward without editing. If you're happy to set up a tool and read its alerts yourself, that's cheaper and you probably don't need us. This is for studios that would rather hand that part over.

We don't build, edit or fix workflows, reconnect apps, or switch anything on or off. We don't hold your API keys or passwords, and we never collect or move money. We're not an around-the-clock service: we check on working days, so a failure on Saturday is flagged on Monday, and we don't call you at night. We can't see inside your clients' apps beyond the run history and the exports you give us. We don't write your client contracts or SLAs, give legal or compliance advice, or work on workflows that handle health information.

Pick one or two client accounts. We read last month's run history and send you the issue log and a sample client report within 5 working days of getting access. Platforms keep only a limited run history, depending on platform and plan, so if there's less than a month we check what's there plus the next two weeks. Then pick a plan or stop. Either way the report and the issue log are yours.

A monthly invoice that you can pay by bank transfer (US ACH or UK bank transfer) or by card.