Bias Correction
Access MeVitae’s API to integrate bias correction into your hiring process
-
Access token
To generate the access token, please follow these steps
As soon as you login to the system. You will be welcomed with this message in the centre of the screen. Please click Generate
You will now have a new access code generated
Please note: The generated token gives access to all the following features and data mentioned below
-
Request Limits
Based on the type of account held by the user the requests are limited to the following for every 10 seconds
Account type Interval Requests Normal Tokens 10 seconds 5 Premium Tokens 10 seconds 20 Enterprise Tokens 10 seconds 50 Users exceeding the limits will receive the following message in their HTTP headers for efficient throttling
Header Description X-RateLimit-Limit The maximum number of requests that the user is allowed to make X-RateLimit-Remaining The number of requests remaining X-RateLimit-Reset The time at which the rate limit resets in seconds. Json result
{
"error": "limit reached, please try back in 4 seconds"
}If you are exceeding the rate limit with these features, please contact us
-
POST
/upload/cvs
For performing just the bias correction please set the job_id to -1. After the upload completes a unique id will be created to retrieve the corrected cvs. this should be a unique
Path Parameters
job_id:
requiredstring
Please set this to be -1 as bias correction does not require job descriptionQuery Parameters
Key Type Description bias_correction string Following are the different types of bias correction MeVitae provides. - none - no tagging will be performed when this option is selected
- redact - All the bias inducing factors get redacted from the CV
- replace - All the bias inducing factors get replaced with machine learnt neutral words in the CV (experimental)
cvs arrays Collection of CVs in the following model Key Type Description cv_name string The name of the cv being uploaded cv_data string CV needs to be in base64 format. MeVitae currently supports docx, pdf or rtf Json return
Uploading the cvs to the /upload/cvs api returns with the following keysKey Type Description wait_time long Estimated wait time for the analysis to complete in seconds upload_date string Returns the uploaded time stamp unique_id string Returns an unique id for the uploaded CVs -
GET
/biascorrected
The user can retrieve all the bias corrected CVs after the wait_time.
Path Parameters
unique_id:
requiredstring
The unique id of the uploaded CVsJson return
Retrieve the collection of all the uploaded bias corrected CVs based on options set during upload. Each CV will have the following keys.Key Type Description cv_name string Name of the CV along with the score e.g cv-82.docx cv_data string Base64 of the CV in Docx format
Support
Please contact [email protected] with any questions.