Relayrix
SFTP automation
Most scheduled SFTP transfers run on something somebody wrote once — a WinSCP script behind a Windows scheduled task, a cron job, a PowerShell file on a server nobody logs into any more. It works, right up until the night it does not.
And the failure you should worry about is not the one that throws an error. It is the file that simply never arrives: nothing failed, so nothing alerted, and the first person to notice is the partner who was waiting for it.
Your SFTP host, with a password or an SSH key. Credentials are stored encrypted, in one place, behind role-based access — not saved in five people’s desktop clients. Test it against the live server before you save.
Pick the source folder and the destination, and set which filenames to include. One pipeline can deliver the same file to more than one destination in a single run.
Every five minutes through every 24 hours, or at a set time on the weekdays you choose, in your own time zone — so 06:00 stays 06:00 through the clock change. A local folder can also be watched, so a transfer fires when a file lands in it.
Every run and every file is recorded, searchable, with a receipt showing where each file went and when it got there. Nobody has to SSH anywhere to answer “did last night’s file go out?”
The one thing a script cannot do for you. Set the deadline you actually work to — “the claims file must be here by 09:00, Monday to Friday” — and a feed that goes quiet raises an alert instead of passing unnoticed.
After each delivery the file is read back off the destination and checked against what was sent. A transfer only counts as delivered once that matches, so “success” never means “we assume it worked.”
Every upload lands under a temporary name and is renamed onto its real one only once every byte is there. A dropped connection leaves them nothing to ingest rather than half a batch they will process anyway.
If the machine is restarted or the container is killed mid-transfer, the interrupted run is reconciled and reported — with how many files made it — rather than sitting on the screen as though it were still going.
Email, or a webhook into Slack, Teams or Discord, on failures and on feeds that go quiet. With a cooldown, so a job failing every few minutes does not send you sixty messages.
Every pipeline has separate Test and Prod environments with their own settings and their own switches. Prove a change works, then promote it — instead of editing the script that is running tonight.
The same pipeline shape covers FTPS, FTP, WebDAV, HTTP(S), Amazon S3, SMB shares and local paths, so a second protocol is not a second tool. HL7, X12 and FHIR can be validated or converted on the way through.
Per-person logins with admin, creator and viewer roles, an audit log nobody can quietly edit, and the whole database encrypted at rest with AES-256.
The same job, in the places it usually turns up.
Overnight lab results out to referring practices, claims to a clearinghouse, remits back in. Feeds that run on a clock, where a file arriving late matters as much as one failing, and HL7 or X12 often needs checking before it goes.
Bank files, positive-pay uploads, lockbox and payroll returns — each on its own cut-off with its own partner. The deadline is the requirement, so an alert that fires when the file has not landed by 09:00 is the point.
Purchase orders, advance ship notices and invoices dropped to a trading partner’s SFTP on a schedule. Dozens of small feeds, each set up by a different person over several years, all of them quietly load-bearing.
Transfers you inherited and now own, spread across clients and servers. One place to run them, one history to search when a client asks, and role-based access so a technician can watch a feed without holding the credentials.
Unlimited pipelines and connections, every connector included, on your own infrastructure. Or start a 14-day free trial — the full product, no credit card.