Start match job
Authorizations
Bearer authentication header of the form Bearer <token>
, where <token>
is your auth token.
Body
name of the job, this must be provided if the job_id is not provided
if you want to run an existing job (e.g. because more records have been added to the dataset), pass the existing job id here
additional records that will be added to the dataset and processed as part of the job
if you want to process a subset of the records in the dataset, pass the batch size here, otherwise all records will be processed
a post request will be sent to this url when the job is completed
email sent to the user triggering the job when the job is completed
custom instructions for the system to follow when completing the job
Existing records that have already been processed by the job will be reprocessed. You may want to do this if you've labeled some more data and want the job to be re-run
If true, the model will return the confidence of the prediction
the dataset id of the dataset whose records will be matched
the dataset id of the dataset whose records will be used to match against the base dataset records. This can be the same as the base dataset id (e.g. deduplication)
the number of potential match records that the model will choose from. A higher number will increase the accuracy of the model but will also increase the cost and time it takes to process each record
if you want to provide labels for the model to use, pass them here. The keys should be the ids of the base dataset records and the values should be the ids of the match dataset records or null if there is no match
Response
match
, transform
, categorize
processing
, completed
, failed
, cancelled