Comparison

Relayrix vs AWS Transfer Family

Transfer Family is how partners get files to you. Relayrix is what happens to those files afterwards — and what tells you when a partner stops sending.

These two products get compared because they both have “transfer” in the description, and because somebody in the room has already paid for one of them. They are not the same kind of thing, and a team that needs both will usually end up running both.

What follows is the honest version: what each one is actually for, where AWS is the better answer, and what an endpoint costs before you count everything it still needs around it.

Two different jobs

What AWS Transfer Family is

A fully managed inbound server endpoint. You enable SFTP, FTPS, FTP or AS2 on it, hand your trading partners credentials, and the files they upload land directly in Amazon S3 or EFS. AWS runs the listener, patches it, and scales it. For that job it is very good, and there is no self-hosted equivalent you would rather operate.

Everything after the file lands is yours to build. The managed workflows that come with it fire only on files uploaded to that same endpoint, and every bucket a workflow touches has to sit in the same region — it is a post-upload hook, not a pipeline engine. Routing the file onward, retrying the one that failed, noticing the partner who went quiet on Tuesday: that is EventBridge, Lambda, Step Functions, CloudWatch and somebody’s afternoon.

What Relayrix is

A client-side automation and monitoring product. It dials out to the systems you already have, moves files between them on a schedule or the moment one lands, and is built to tell you when something did not happen.

What you get that an endpoint does not give you

It routes anywhere, not into one vendor’s storage

SFTP to an SMB share, FTPS to S3, a database query to a partner’s SFTP, a watched folder to three destinations at once. Nothing has to touch S3, and nothing has to be in the same region as anything else. Eight connectors — SFTP, FTPS, FTP, WebDAV, HTTP(S), S3 and S3-compatible, SMB/CIFS and local paths — plus SQL Server, MySQL, PostgreSQL and MongoDB as query sources.

It tells you when a file never arrives

The failure worth worrying about is not the one that throws an error — it is the file that simply never shows up. Nothing failed, so nothing alerted, and the first person to notice is the partner who was waiting. 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 on its own.

Every failed file has somewhere to be

A file that did not go through stays on a Needs attention list until it does or somebody clears it, with the reason in plain language rather than a stack trace. Individual files retry on their own; a run that was cut off mid-transfer is reconciled and reported rather than left looking as though it were still going.

Connection health, from real traffic

Which server is actually down right now, and how many pipelines that affects — measured from the transfers that just ran, not from a synthetic ping.

A delivery receipt for every file

After delivery the file is read back off the destination and checked against what was sent, so “success” never means “we assume it worked.” Uploads land under a temporary name and are renamed onto the real one only once every byte is there, so a partner never picks up half a file.

Reports that arrive without being asked for

Scheduled email summaries that count records, not just files, with calendar and rolling date ranges and period-over-period comparison. This is usually what the manager signing the invoice actually wanted.

A screen your ops staff can run

Pipelines are built in a browser by the person who owns the feed, with separate Test and Prod settings so a change can be proved before it is live. Nobody needs an AWS console login, an IAM policy or a deployment to change a folder name.

It runs where you do

On your own hardware, in your own network, air-gapped if that is the requirement — Docker for Linux and macOS, or a native Windows Server installer backed by your own SQL Server. Nothing phones home, and there is no telemetry to switch off. Built for HIPAA compliance, with AES-256 at rest, role-based access and a tamper-evident audit log. A buyer who cannot egress to a hyperscaler at all still has a product here.

When AWS Transfer Family is the right answer

Four cases where we would tell you to buy theirs. If your requirement is on this list, the rest of this page is not the argument you need.

Partners need to log in to you

If the job is “give forty trading partners their own SFTP credentials against an endpoint we host,” Transfer Family does exactly that and does it well. Relayrix connections dial out to the partner’s server, which is the opposite direction. Buy theirs.
The exchange has to be AS2

Retail and some claims trading partners mandate AS2 with signed receipts, and there is no substituting a different protocol when a partner agreement names one. Transfer Family speaks it natively and bills per message.
You need AWS-grade availability, under an AWS agreement

A multi-AZ managed endpoint with an AWS service-level agreement behind it, covered by the AWS BAA your legal team has already signed, is a real thing to want. We are a single-node deployment you run yourself, and we say so on the pricing page.
The files start and finish inside AWS

If every producer and every consumer of the file is already an AWS service in the same region, the glue between them is short and the managed-workflow hook may be all you need. The case for a separate product gets stronger the moment one end of the route is not in AWS.

What an endpoint costs

Transfer Family bills $0.30 per hour for each protocol you enable, whether or not anyone connects. One protocol, running continuously, is about $219 a month — roughly $2,630 a year. Two protocols is two of those. On top of that, $0.04 per GB uploaded and downloaded.

Then the parts that are not on the Transfer Family line of the bill: the S3 storage the files land in, the CloudWatch logs you will need to answer “did last night’s file arrive,” the Lambda and Step Functions that move the file anywhere other than that bucket, Secrets Manager, and NAT or Route 53 if the design needs them. Separately metered again if you use them: AS2 at $0.01 per message for the first 100,000 a month, SFTP Connectors at $0.001 per call plus $0.40 per GB, Web Apps at $0.50 per hour per 250-session unit, and PGP decrypt workflow steps at $0.10 per GB.

None of which is a criticism of the price — it is a fair price for a managed listener. The point is what it buys: an endpoint. Not pipelines, not alerting on the file that never came, not a delivery receipt, not a report, and not a screen anyone outside the platform team can use. Those are the things you would still be building afterwards, and they are what we sell.

Figures taken from the AWS Transfer Family pricing page and verified September 2026. AWS changes pricing; check it against the source before quoting it.

Most teams run both

The common shape is Transfer Family as the front door for partners who need to push files to you, and Relayrix behind it: pulling from the S3 bucket those files land in, validating the HL7 or X12 inside them, routing each one to the system that needs it, decrypting or encrypting on the way through, and raising an alert when a partner who sends every weekday morning does not.

The same install handles the feeds that never touch AWS at all — the overnight pull from a payer’s SFTP, the drop onto a shared drive, the extract out of SQL Server — so one team watches one screen instead of two.

See what happens after the file lands

One flat annual license, self-hosted, every connector and every feature included. Or start a 14-day free trial of the full product — no credit card.

Start a free trial See pricing