Quickstart
This guide walks you through collecting a vendor file, validating it, and delivering a standardized version.
Before you start
You need a VendorGate account with access to the application. If you do not have one, contact your account administrator or request early access.
Step 1: Add a credential
Most vendor sources require authentication. Start by creating a credential that VendorGate can use to reach your source.
- Go to Credentials in the application.
- Click New Credential.
- Choose the credential type that matches your source, for example SFTP or HTTP API.
- Enter the required details and save.
Step 2: Create a contract
A contract tells VendorGate what a valid file looks like.
- Go to Contracts.
- Click New Contract.
- Add the columns you expect, such as
supplier_id,invoice_date, andamount. - For each column, set the expected data type and whether it is required.
- Add any rules, such as
amount > 0.
Step 3: Build a pipeline
A pipeline pulls the file, validates it, and sends it forward.
- Go to Pipelines.
- Click New Pipeline.
- Select your source type and the credential you created in step 1.
- Choose the contract you created in step 2.
- Choose a delivery destination, such as SFTP, FTP, S3, an HTTP API, a database, or a lookup dataset.
- Save and trigger a test run.
Step 4: Add a transformation
If the validated file needs to be cleaned or reshaped, add a standardizer.
- Go to Transformations.
- Click New Standardizer.
- Select the pipeline output as the input.
- Define the output columns you want.
- Map input fields to output columns. Choose a mapping mode for each column: simple field mapping, multi-step transforms, expressions, or LLM-powered mapping.
- Save and run the standardizer.
What happens next
When the standardizer runs successfully, the output file is delivered to the destination you configured. You can view run history, inspect errors, and refine mappings over time.
Tip
Start with a small sample file when testing a new pipeline or standardizer. This makes it easier to spot contract or mapping issues before processing production volumes.