Scoring API now returns job scores built for candidates

Scoring algorithms built on the Hyperion "AI Job board" model now return job scores built for candidates: each score reflects how well the job fits the candidate, so it can be shown as is to the candidates you recommend jobs to.

Action required

Now: none. GET /jobs/scoring and GET /profiles/scoring keep their parameters and their response format.
Later: none. Nothing in this update is deprecated.

What changed

Scores of algorithms built on the Hyperion "AI Job board" model are now built for candidates: a job that fits the candidate well scores high and one that fits poorly scores low, so each score can be read on its own. Before, scores sat close together, so they were best read by comparing results with each other. For example, jobs that scored 0.93, 0.90 and 0.85 can now score 0.85, 0.25 and 0.02.

Example

curl -G https://api.hrflow.ai/v1/jobs/scoring \
  -H "X-API-KEY: <your_api_key>" \
  -H "X-USER-EMAIL: <your_email>" \
  --data-urlencode "algorithm_key=<your_algorithm_key>" \
  --data-urlencode "source_key=<your_source_key>" \
  --data-urlencode "profile_key=<your_profile_key>" \
  --data-urlencode 'board_keys=["<your_board_key>"]'

The response keeps its shape: data.jobs and data.predictions, one [1-score, score] pair per job.

{
  "data": {
    "jobs": [{"key": "<job_key>"}],
    "predictions": [[0.49184858798980713, 0.50815147161483765]]
  }
}

Action required

What changed

Example

Links