How do I request a new field or data model?
Updated 2 weeks ago
Check raw_data first: if the value is there, you can map it yourself with a custom field. If it isn't, tell us the field, the provider and your use case, and we add it or suggest another route.
Check whether the value already reaches us
Call the endpoint with include_raw_data=true and look for the value in the provider's original payload.
What you see | What to do |
|---|---|
The value is in | Create a custom field that reads it, then request it with |
The value isn't in | Ask us to add it. We add it to |
The value isn't in the provider's API at all | We can't sync it. Some data only exists in the provider's interface or in reports. |
A value can also be missing because the connection's scope or permissions exclude it. See Scoping to tell a permission gap from a mapping gap.
Request a new data model
When you need a whole concept the unified models don't cover, send us the use case, the providers involved, and an example of the output you expect. We draft the model with you and confirm it covers your cases before we build it.
For providers where the unified models cover only part of the data, Passthrough lets you call the provider's own endpoints through the connection you already have. PrismHR is a common example, since much of its data exchange runs through file uploads that don't map cleanly to a unified schema.
What limits what we can add
The provider's API. If the API doesn't expose a value, we can't return it. For example, the 7shifts API has no field that marks a user's primary location or role.
The customer's own data. Some systems, such as PrismHR, return only what the employer or PEO keeps in them. An empty field in the source stays empty in our response.