Candidate Search

Search for strong candidates for open jobs.

Overview


Candidate Search enables partner Applicant Tracking Systems and other HR Tech vendors to power features that recommend best-fit candidates for a job. The partner provides candidates.

End users search for candidates by detailing a target Job which describes the attributes of an ideal candidate (preferred skills, education, and more), and filter(s) which restrict the candidate pool (on location, education, and more). End users are presented with candidates who have the skills, work experience, and education to be successful in the job.

For each recommended candidate, AdeptID provides a Match consisting of assessments of work experience, skills, and education; a summary assessment, a skills report describing skill gaps and overlaps, and a narrative explanation of the match quality. The full details of the match help end users understand the "why" behind each recommendation.

Partners provide candidates through one or more private candidate feeds. The feed defines the pool of candidates for a given candidate search. The feed contents are independent of each other, fully managed by the partner. A candidate search targets a single candidate feed.

For more information on how our modeling works, head to the Modeling Approach section.


API Endpoints

Model Inputs and Outputs

Inputs

Outputs

For each Job:

  • Job title, employer, skills, education, job description.

Filter criteria:

  • Location, required education level, required skills, etc.

Recommended Candidates: List of recommended Candidates and their corresponding Matches .

Match Assessments :

  • Summary match quality ranking (Low to Very High) and match score for each candidate.
  • Attribute scores assessing a candidate's fit with respect to skills, work experience and education. (available through private beta)
  • Skills Report listing the top skill overlaps and gaps for each candidate, in relation to the job.
  • Narrative summarizing the overall match in 2-4 sentences. (available through private beta)